| Objective:The Corona Virus Disease 19(COVID-19)epidemic has caused serious harm to human health,and social and economic development,and posed a serious threat to the field of public health.The development trend of the epidemic is directly related to the health of people all over the country and even the world.The effective containment of the epidemic depends on the understanding of the risk and future development trend of the epidemic,the rapid acquisition of on-the-spot information of epidemiological investigation,and the adoption of timely and effective prevention and control measures.Epidemic prevention and control need to obtain information based on epidemiological investigation to determine the cause,source of infection,route of transmission,susceptible population and risk factors of,the disease,and to predict the epidemic trend.Therefore,it is necessary to construct prediction models for different scenarios of novel coronavirus infection and explore different epidemiological investigation methods.Based on the actual situation of epidemic prevention and control,this study provides an effective tool for the formulation of prevention and control countermeasures against novel coronavirus infection or sudden infectious diseases by relying on infectious disease models.Relying on the advanced software technology to obtain the epidemiological investigation field information to provide a convenient and fast tool.Methods:The 570 laboratory-confirmed cases in Yangzhou from 28 July to 26 August 2021 collected in this study were obtained from the National Health Commission of the People’s Republic of China(NHC)and the local Health Commission.This information includes demographic characteristics(age,sex,residence,and medical treatment),contact history,close contacts,clinical type,critical time points(nucleic acid test date,isolation date,diagnosis date,etc.),activity paths,and chess and card room location,etc.Now,all this information is used to construct an improved Susceptible-exposeinfected-Recovered(SEIR)model and an intelligent-based model and to simulate the effects of different control policies in different scenarios.Case Report Form(CRF)technology,human-computer interaction(HCI)and decision recommendation framework were used to construct the customized system of epidemiological questionnaires.Epidemiological investigators could customize and issue questionnaires according to different needs.Respondents could fill in epidemiological investigation information and upload documents.The data of the cough voice print detection model was collected by communication intelligent robot human-computer voice interaction technology,including 605 cough voice records of patients(clinically diagnosed with respiratory diseases)and healthy people(non-respiratory diseases).The training validation of the Convolutional Recurrent Neural Network(CRNN)model is used as a screening tool for respiratory diseases.Results:1.Simulation of different epidemic scenarios in YangzhouA total of 570 people were infected in this epidemic.The simulation results of the large-scale epidemic situation in Yangzhou City by the improved SEIR model showed that for people from medium-high risk areas,nucleic acid detection within 3 days before travel could limit the total number of infected people in Yangzhou to 50.Among the elderly,the total number of infected people in Yangzhou will decrease by 119 when the vaccination rate for the elderly is raised to 30 percent,and 253 when the vaccination rate for people over 60 years old is raised to 60 percent.The simulation result of the agent-based model for the small scene represented by the chess and card room shows that 75 people will be infected in the chess and card room.When the population density of the chess and card room decreases by 40%,the number of infected people will decrease by 54 in 7 days.When the ventilation in the card room was increased from25% to 50%,the total number of infections decreased by 33 in 7 days;When ventilation was increased from 25 percent to 75 percent,the total number of infections decreased by 63 in seven days.2.Epidemiological questionnaire customization systemThis study realized the function of customizing the epidemiological questionnaire,and the managers managed the existing questionnaire through the system.The investigated person fills out the questionnaire and uploads the document through the link or two-dimensional code.Epidemiological investigators can customize and release the questionnaire according to the needs of different investigation scenarios such as diseases and events,and browse,modify,delete,query,analyze and export the survey data3.Respiratory disease screening moduleThe accuracy of the simulation quality of respiratory diseases was 96.61%,and the accuracy rate,recall rate,and F1 score were all 90%.By comparing the simulation results with the "gold standard" of clinical diagnosis,the sensitivity and specificity were 90% and 97.95%,respectively.Conclusion:In this study,the simulation model of novel coronavirus infection and the customized epidemiological questionnaire system for different scenarios are constructed,which can realize the screening at the early stage of the disease,the epidemiological investigation in the development process,and the simulation of the epidemic trend,and can provide a powerful tool for the prevention and control of infectious diseases.1.The simulation of epidemic prevention and control measures for large scenarios showed that the shorter the interval between nucleic acid test time and travel date,the better the detection effect.It is recommended to conduct nucleic acid test one day before travel.2.In the early stage of the epidemic in Yangzhou,the elderly were the majority.It is suggested to expand the age restriction of the vaccine strategy on the premise of ensuring safety,and to advocate the vaccination of healthy elderly people.3.The simulation of epidemic prevention and control measures in small scenes suggests that ventilation volume and social distancing should be increased in small scenes.4.The customized epidemiological questionnaire system can realize the smooth transmission of the front and rear data so that the rear experts can understand the actual situation of the epidemiological investigation site more intuitively through the system,and carry out timely and accurate guidance to the investigation work carried out by the front epidemiologist investigators,so as to realize the effective development of the field investigation and epidemic treatment.5.The screening model of respiratory diseases has high accuracy,which can realize the identification of lung diseases through training,and provide diagnostic ideas for the early screening of respiratory diseases and the possible emergence of sudden respiratory infectious diseases. |
PDF Full Text Request