A public health emergency system management method and system based on model feedback

Abstract

The present invention relates to the field of public health emergency management, and discloses a public health emergency system management method and system based on model feedback, including the following steps: collecting various historical indicator data of the public health emergency system for constructing a vulnerability measurement model of the public health emergency system, and evaluating the vulnerability status of the public health emergency system through the vulnerability measurement model of the public health emergency system, and at the same time, formulating intervention strategies and successively outputting interventions for areas with abnormal vulnerability status. The present invention can analyze the vulnerability of public health events in primary medical institutions by constructing a vulnerability measurement model of the public health emergency system, and propose intervention strategies, thereby providing ideas and countermeasures for improving the ability and level of primary medical institutions to respond to public health emergencies, and reducing casualties in public health events.

Description

Technical Field

[0001] The present invention relates to the field of public health emergency management, and in particular to a public health emergency system management method and system based on model feedback. Background Art

[0002] The public health emergency response system refers to a system that responds and manages quickly and efficiently through comprehensive coordination of organizations, mechanisms, resources and measures to protect public health, reduce losses and restore normal social operations when responding to public health emergencies, such as infectious disease outbreaks, natural disasters, chemical accidents, etc. The system involves coordination and cooperation at multiple levels, including government departments, medical institutions, emergency response teams, communities and the public.

[0003] As the frontline for handling public health emergencies, the emergency management capabilities of primary medical institutions directly determine the bottom line of a country's overall emergency management work level. Building an efficient public health emergency system and management system for primary medical institutions and promoting the improvement of public health emergency management level have become an important task in the country's economic and social construction.

[0004] This study takes the public health emergency response system of primary medical institutions as the research object, adopts the grounded theory analysis method, collects data through interviews, constructs a vulnerability measurement model of the public health emergency response system of primary medical institutions with three dimensions of "exposure, sensitivity, and adaptability", applies field surveys to collect data, and uses scientific statistical analysis methods to analyze the current status of the vulnerability of the public health emergency response system of primary medical institutions and its influencing factors, evaluates the vulnerability of the public health emergency response system of primary medical institutions, and proposes intervention strategies, thereby providing ideas and countermeasures for improving the ability and level of primary medical institutions to respond to public health emergencies. Summary of the invention

[0005] The present invention overcomes the deficiencies of the prior art and provides a public health emergency system management method and system based on model feedback.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is:

[0007] The first aspect of the present invention provides a public health emergency system management method based on model feedback, comprising the following steps:

[0008] Collect historical indicator data of the public health emergency system, and build a vulnerability measurement model for the public health emergency system based on the indicator data of the public health emergency system;

[0009] Based on the vulnerability measurement model of the public health emergency system, obtain the functional score of the current indicator data of the target public health emergency system, and conduct a vulnerability assessment of the target public health emergency system based on the functional score of the current indicator data;

[0010] Formulate intervention strategies for vulnerable abnormal areas, and output prioritized interventions for vulnerable abnormal areas based on the ranking table of total functional scores of vulnerable abnormal areas.

[0011] Furthermore, in a preferred embodiment of the present invention, the historical indicator data of the public health emergency system is collected, and a vulnerability measurement model of the public health emergency system is constructed based on the indicator data of the public health emergency system, specifically:

[0012] The primary medical institutions that need to be managed by the public health emergency system are identified as target primary medical institutions, and the current public health emergency system in the target primary medical institutions is identified as the target public health emergency system;

[0013] Acquire a historical data network, and acquire historical indicator data of a target public health emergency system through the historical data network, wherein the historical indicator data includes public health data, health infrastructure data, emergency management data of a target primary medical institution, and population and socioeconomic data of a location of the target primary medical institution;

[0014] Based on the historical indicator data, construct a public health emergency survey questionnaire, and publish the public health emergency survey questionnaire to target primary medical institutions for indicator survey;

[0015] The principal component analysis method is introduced, and based on the principal component analysis method, the data on the public health emergency survey questionnaire after the indicator survey is converted into a feature matrix, which is calibrated as a type of feature matrix;

[0016] Performing eigenvalue decomposition on the one type of characteristic matrix to obtain a type of eigenvector value, and sorting all the one type of eigenvector values ​​by eigenvalue to obtain a sorting table of the one type of eigenvector values;

[0017] Based on the first type of feature vector value sorting table, weights of different historical indicator data are calculated, and according to the weights of the different historical indicator data, a historical indicator data weight sorting table is constructed;

[0018] A long short-term memory network algorithm is introduced, and a vulnerability measurement model for the public health emergency system is constructed based on the long short-term memory network algorithm and a weight ranking table of historical indicator data.

[0019] Furthermore, in a preferred embodiment of the present invention, the long short-term memory network algorithm is introduced, and a vulnerability measurement model of the public health emergency system is constructed based on the long short-term memory network algorithm and the weight ranking table of historical indicator data, specifically:

[0020] Presetting a public health emergency system vulnerability prediction time period, and constructing a public health emergency system vulnerability prediction timestamp based on the public health emergency system vulnerability prediction time period;

[0021] Combining the vulnerability prediction timestamp of the public health emergency system and the historical indicator data, constructing historical indicator data in a time series format, calibrating it as historical indicator time series data, and updating the historical indicator time series data based on the historical indicator data weight ranking table to obtain historical indicator weight ranking time series data;

[0022] Introducing a long short-term memory network algorithm, constructing a long short-term memory network basic model, and obtaining an input layer, an LSTM layer, and a fully connected layer in the long short-term memory network basic model;

[0023] Based on the weighted sorting time series data of the historical indicators, the input dimension data of the input layer is determined and calibrated as the target input dimension data;

[0024] Based on the historical data network, when retrieving and building the long short-term memory network model, the standard number of LSTM layers and the standard number of neurons in a single LSTM layer corresponding to different data volumes;

[0025] The data volume in the weighted time series data of the historical indicators is obtained and calibrated as the target data volume. Based on the target data volume, the target number of LSTM layers and the target number of neurons in a single LSTM layer are obtained by combining the standard number of LSTM layers and the standard number of neurons in a single LSTM layer corresponding to different data volumes.

[0026] Based on the target input dimension data, the target number of LSTM layers, and the target number of neurons in a single LSTM layer, the target input layer and the target LSTM layer are constructed. The target input layer, the target LSTM layer, and the single fully connected layer are combined to construct a long short-term memory network training model. The historical indicator weight-sorted time series data is imported into the long short-term memory network training model to obtain a vulnerability measurement model for the public health emergency system.

[0027] Furthermore, in a preferred embodiment of the present invention, the vulnerability measurement model of the public health emergency system is used to obtain the functional score of the current status indicator data of the target public health emergency system, and the vulnerability assessment of the target public health emergency system is performed based on the functional score of the current status indicator data, specifically:

[0028] In the target primary medical institution, the current status indicator data of the target public health emergency system is obtained, the vulnerability measurement model of the public health emergency system is run, and the current status indicator data of the target public health emergency system is imported into the vulnerability measurement model of the public health emergency system for storage;

[0029] In the vulnerability measurement model of the public health emergency response system, a function score qualification threshold is preset, and the function score qualification threshold includes an exposure score qualification threshold, a sensitivity score qualification threshold, and a fitness score qualification threshold;

[0030] In the vulnerability measurement model of the public health emergency system, a vulnerability assessment is performed on the stored status indicator data of the target public health emergency system to obtain a functional score of the status indicator data of the target public health emergency system, wherein the functional score includes an exposure score, a sensitivity score, and a fitness score;

[0031] If the function scores of the current status indicator data of the target public health emergency system are all within the qualified function score threshold, the target public health emergency system will be marked as a qualified public health emergency system;

[0032] If any score in the functional score of the current status indicator data of the target public health emergency system is not within the functional score qualification threshold, the target public health emergency system will be marked as a public health emergency system with abnormal current status, and the public health emergency system with abnormal current status will be divided into regions and the vulnerability assessment of different regions will be ranked.

[0033] Furthermore, in a preferred embodiment of the present invention, the regional division of the abnormal status public health emergency response system and the vulnerability assessment ranking of different regions are specifically as follows:

[0034] Based on the current abnormal public health emergency response system and the public health emergency functionality, the target primary medical institutions are divided into different areas and marked as vulnerability areas to be assessed;

[0035] Obtaining a functional score of the status indicator data of the current abnormal public health emergency response system, and based on the functional score of the status indicator data of the current abnormal public health emergency response system, marking the vulnerability to be assessed area whose functional score of the status indicator data is maintained within the functional score qualified threshold as a vulnerability qualified area in the vulnerability to be assessed area;

[0036] In the vulnerability assessment area, the vulnerability assessment area where the function score of the status indicator data does not remain within the qualified function score threshold is marked as an abnormal vulnerability area;

[0037] The functional scores of different status indicator data in the vulnerability anomaly area are added together to obtain the total functional score of the status indicator data. The total functional scores of the status indicator data of all vulnerability anomaly areas are compared, and a ranking table of the total functional scores of vulnerability anomaly areas is constructed based on the comparison results.

[0038] Furthermore, in a preferred embodiment of the present invention, the intervention strategy for the abnormal vulnerability area is formulated, and based on the ranking table of the total functional scores of the abnormal vulnerability areas, the abnormal vulnerability areas are intervened in sequence and output, specifically:

[0039] The functional scores of the status quo indicator data in the vulnerability anomaly area are analyzed, and based on the analysis results, the vulnerability anomaly area is divided into exposure anomaly area, sensitivity anomaly area and fitness anomaly area;

[0040] For the abnormal exposure area, obtain the current indicator data of the abnormal exposure area, and retrieve the intervention strategy of the abnormal exposure area based on the current indicator data of the abnormal exposure area in the historical data network;

[0041] For the abnormal sensitivity area, obtain the current status indicator data of the abnormal sensitivity area, and retrieve the intervention strategy of the abnormal sensitivity area based on the current status indicator data of the abnormal sensitivity area in the historical data network;

[0042] For the fitness abnormal area, obtain the current indicator data of the fitness abnormal area, and retrieve the intervention strategy of the fitness abnormal area based on the current indicator data of the fitness abnormal area in the historical data network;

[0043] The intervention strategies for the exposure abnormality area, the sensitivity abnormality area and the fitness abnormality area are integrated to obtain the total intervention strategy for the vulnerability abnormality area, and the total intervention strategy for the vulnerability abnormality area is output to different vulnerability abnormality areas in sequence based on the ascending principle in the vulnerability abnormality area function total score ranking table.

[0044] The second aspect of the present invention further provides a public health emergency system management system based on model feedback, the public health emergency system management system includes a memory and a processor, the memory stores a public health emergency system management method, and when the public health emergency system management method is executed by the processor, the following steps are implemented:

[0045] Collect historical indicator data of the public health emergency system, and build a vulnerability measurement model for the public health emergency system based on the indicator data of the public health emergency system;

[0046] Based on the vulnerability measurement model of the public health emergency system, obtain the functional score of the current indicator data of the target public health emergency system, and conduct a vulnerability assessment of the target public health emergency system based on the functional score of the current indicator data;

[0047] Formulate intervention strategies for vulnerable abnormal areas, and output prioritized interventions for vulnerable abnormal areas based on the ranking table of total functional scores of vulnerable abnormal areas.

[0048] The present invention solves the technical defects existing in the background technology, and the present invention has the following beneficial effects: various historical indicator data of the public health emergency system are collected to construct a vulnerability measurement model of the public health emergency system, and the vulnerability status of the public health emergency system is evaluated through the vulnerability measurement model of the public health emergency system, while the intervention strategy is formulated and the intervention output is carried out successively for the areas with abnormal vulnerability status. The present invention can analyze the vulnerability of public health events in primary medical institutions by constructing a vulnerability measurement model of the public health emergency system, and propose intervention strategies, thereby providing ideas and countermeasures for improving the ability and level of primary medical institutions to respond to public health emergencies, and reducing casualties in public health events. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, drawings of other embodiments can be obtained based on these drawings without paying creative work.

[0050] Figure 1 A flow chart showing a method for managing a public health emergency system based on model feedback is shown;

[0051] Figure 2 A flow chart of the method for constructing a vulnerability measurement model for a public health emergency system is shown;

[0052] Figure 3 A procedural view of a public health emergency system management system based on model feedback is shown. DETAILED DESCRIPTION

[0053] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0054] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0055] Figure 1 A flowchart of a public health emergency system management method based on model feedback is shown, comprising the following steps:

[0056] S102: Collect historical indicator data of the public health emergency system, and build a vulnerability measurement model of the public health emergency system based on the indicator data of the public health emergency system;

[0057] S104: Based on the vulnerability measurement model of the public health emergency system, obtain the functional score of the current status indicator data of the target public health emergency system, and conduct a vulnerability assessment on the target public health emergency system based on the functional score of the current status indicator data;

[0058] S106: Formulate intervention strategies for the abnormal vulnerability areas, and output the priority intervention for the abnormal vulnerability areas based on the ranking table of the total functional scores of the abnormal vulnerability areas.

[0059] Furthermore, in a preferred embodiment of the present invention, the vulnerability measurement model of the public health emergency system is used to obtain the functional score of the current status indicator data of the target public health emergency system, and the vulnerability assessment of the target public health emergency system is performed based on the functional score of the current status indicator data, specifically:

[0060] In the target primary medical institution, the current status indicator data of the target public health emergency system is obtained, the vulnerability measurement model of the public health emergency system is run, and the current status indicator data of the target public health emergency system is imported into the vulnerability measurement model of the public health emergency system for storage;

[0061] In the vulnerability measurement model of the public health emergency response system, a function score qualification threshold is preset, and the function score qualification threshold includes an exposure score qualification threshold, a sensitivity score qualification threshold, and a fitness score qualification threshold;

[0062] In the vulnerability measurement model of the public health emergency system, a vulnerability assessment is performed on the stored status indicator data of the target public health emergency system to obtain a functional score of the status indicator data of the target public health emergency system, wherein the functional score includes an exposure score, a sensitivity score, and a fitness score;

[0063] If the function scores of the current status indicator data of the target public health emergency system are all within the qualified function score threshold, the target public health emergency system will be marked as a qualified public health emergency system;

[0064] If any score in the functional score of the current status indicator data of the target public health emergency system is not within the functional score qualification threshold, the target public health emergency system will be marked as a public health emergency system with abnormal current status, and the public health emergency system with abnormal current status will be divided into regions and the vulnerability assessment of different regions will be ranked.

[0065] It should be noted that the vulnerability measurement model of the public health emergency system can predict the vulnerability of the current situation in the target public health emergency system. For example, the vulnerability measurement model of the public health emergency system can predict the range of infectious disease transmission speed in the target primary medical institutions, or predict the outbreak time of potential epidemics, changes in the supply and demand of medical equipment and medical drugs, etc. Therefore, the current status indicator data of the target public health emergency system is imported into the vulnerability measurement model of the public health emergency system for analysis, and in the analysis process, the exposure, sensitivity and adaptability of the current status indicator data of the target public health emergency system are mainly analyzed. Among them, the exposure is the risk level of the medical structure facing public health events, including natural disasters, infectious disease outbreaks, etc.; the sensitivity is the vulnerability level of the medical institution in public health events, such as the number of medical staff, medical resource reserves, etc.; the adaptability is the ability of the medical institution to respond to and recover from public health events, such as emergency plans, training exercises, community participation, etc. The exposure, sensitivity and adaptability of the current status indicator data of the target public health system are scored, and the public health emergency system is classified based on the scoring results to obtain the current status abnormal public health emergency system. The current abnormal public health emergency response system is an unqualified public health emergency response system. If the public health emergency response system of a medical institution is an abnormal public health emergency response system, it is easy to increase the probability of medical accidents, so it is necessary to intervene and adjust the current abnormal public health emergency response system.

[0066] Furthermore, in a preferred embodiment of the present invention, the regional division of the abnormal status public health emergency response system and the vulnerability assessment ranking of different regions are specifically as follows:

[0067] Based on the current abnormal public health emergency response system and the public health emergency functionality, the target primary medical institutions are divided into different areas and marked as vulnerability areas to be assessed;

[0068] Obtaining a functional score of the status indicator data of the current abnormal public health emergency response system, and based on the functional score of the status indicator data of the current abnormal public health emergency response system, marking the vulnerability to be assessed area whose functional score of the status indicator data is maintained within the functional score qualified threshold as a vulnerability qualified area in the vulnerability to be assessed area;

[0069] In the vulnerability assessment area, the vulnerability assessment area where the function score of the status indicator data does not remain within the qualified function score threshold is marked as an abnormal vulnerability area;

[0070] The functional scores of different status indicator data in the vulnerability anomaly area are added together to obtain the total functional score of the status indicator data. The total functional scores of the status indicator data of all vulnerability anomaly areas are compared, and a ranking table of the total functional scores of vulnerability anomaly areas is constructed based on the comparison results.

[0071] It should be noted that in the target primary medical institutions in the current abnormal public health emergency system, not all parts of the functional scores are unqualified. For example, in the primary medical institutions, the supply and demand of medical equipment and medical drugs may be good, but the number of medical staff cannot meet expectations. Therefore, in the target primary medical institutions, the target primary medical institutions are divided into different areas according to the public health emergency functionality. The public health emergency functionality is the function that can be realized by different parts of the target primary medical institutions. After obtaining the vulnerability abnormal area, the functional score of its current indicator data is analyzed. When the functional score of the current indicator data is within the preset threshold, that is, the exposure score, sensitivity score and fitness score are all within the corresponding threshold, it proves that the public health emergency function of the part of the area is qualified and does not need to be adjusted and managed. On the contrary, if the functional score of the current indicator data of the vulnerability abnormal area is not within the preset threshold, it proves that its exposure score, sensitivity score and fitness score are abnormal. At this time, the area needs to be intervened and improved, and it is marked as a vulnerability abnormal area. There may be many vulnerability abnormal areas, and the order of intervention and improvement needs to be determined. The more serious the abnormal situation, the priority intervention and improvement is required. The order of intervention improvement can be determined by adding up all functional scores in the vulnerability abnormality area and constructing a ranking table of the total functional scores in the vulnerability abnormality area based on the total score.

[0072] Furthermore, in a preferred embodiment of the present invention, the intervention strategy for the abnormal vulnerability area is formulated, and based on the ranking table of the total functional scores of the abnormal vulnerability areas, the abnormal vulnerability areas are intervened in sequence and output, specifically:

[0073] The functional scores of the status quo indicator data in the vulnerability anomaly area are analyzed, and based on the analysis results, the vulnerability anomaly area is divided into exposure anomaly area, sensitivity anomaly area and fitness anomaly area;

[0074] For the abnormal exposure area, obtain the current indicator data of the abnormal exposure area, and retrieve the intervention strategy of the abnormal exposure area based on the current indicator data of the abnormal exposure area in the historical data network;

[0075] For the abnormal sensitivity area, obtain the current status indicator data of the abnormal sensitivity area, and retrieve the intervention strategy of the abnormal sensitivity area based on the current status indicator data of the abnormal sensitivity area in the historical data network;

[0076] For the fitness abnormal area, obtain the current indicator data of the fitness abnormal area, and retrieve the intervention strategy of the fitness abnormal area based on the current indicator data of the fitness abnormal area in the historical data network;

[0077] The intervention strategies for the exposure abnormality area, the sensitivity abnormality area and the fitness abnormality area are integrated to obtain the total intervention strategy for the vulnerability abnormality area, and the total intervention strategy for the vulnerability abnormality area is output to different vulnerability abnormality areas in sequence based on the ascending principle in the vulnerability abnormality area function total score ranking table.

[0078] It should be noted that different vulnerability anomaly areas require different intervention strategies, that is, different intervention strategies are required for exposure anomaly areas, sensitivity anomaly areas and fitness anomaly areas. The historical data network is a data network that stores various intervention strategies, which is equivalent to a database. Among them, the intervention strategies for exposure anomaly areas include but are not limited to improving the accuracy and coverage of natural disaster warning systems, strengthening the construction of disaster protection facilities, strengthening disease monitoring and epidemiological investigations, promoting vaccination, and improving public health publicity. Etc.; the intervention strategies for sensitivity anomaly areas include but are not limited to increasing the investment in medical staff and medical equipment, optimizing resource allocation, ensuring sufficient medical reserves at critical moments, expanding hospitals and emergency centers, and improving the service capabilities of medical institutions and their ability to respond to emergencies; the intervention strategies for fitness anomaly areas include but are not limited to regularly updating and improving public health emergency plans, ensuring the operability and comprehensive coverage of emergency plans, strengthening emergency training and drills for medical staff and related staff, and improving practical operation capabilities. After combining all intervention strategies to obtain the total intervention strategy for the vulnerability anomaly area, the total intervention strategy for the vulnerability anomaly area needs to be applied to the vulnerability anomaly area, and when applying it, it needs to be sorted based on the ranking table of the total function score of the vulnerability anomaly area. For example, areas with lower total scores need more limited intervention, so the principle of ascending order is continued to output the total intervention strategy for vulnerability anomaly areas to different vulnerability anomaly areas one after another.

[0079] Figure 2 A flowchart of a method for constructing a vulnerability measurement model for a public health emergency system is shown, comprising the following steps:

[0080] S202: Obtain target primary medical institutions, target public health emergency response systems, and historical indicator data of target public health emergency response systems;

[0081] S204: Based on the principal component analysis algorithm, weight analysis is performed on the data on the public health emergency survey questionnaire after the indicator survey to obtain a weight ranking table of historical indicator data;

[0082] S206: Introduce a long short-term memory network algorithm, and construct a vulnerability measurement model for the public health emergency system based on the long short-term memory network algorithm and a weight ranking table of historical indicator data.

[0083] Furthermore, in a preferred embodiment of the present invention, the acquisition of the target primary medical institution, the target public health emergency system and the historical indicator data of the target public health emergency system is specifically:

[0084] The primary medical institutions that need to be managed by the public health emergency system are identified as target primary medical institutions, and the current public health emergency system in the target primary medical institutions is identified as the target public health emergency system;

[0085] Acquire a historical data network, and acquire historical indicator data of a target public health emergency system through the historical data network, wherein the historical indicator data includes public health data, health infrastructure data, emergency management data of a target primary medical institution, and population and socioeconomic data of a location of the target primary medical institution;

[0086] Based on the historical indicator data, a public health emergency survey questionnaire is constructed, and the public health emergency survey questionnaire is released to target primary medical institutions for indicator survey.

[0087] It should be noted that the public health emergency system refers to a complex of organizations, mechanisms and resources specifically used to respond to and manage public health emergencies. In primary medical institutions, if an accident such as an infectious disease outbreak occurs, the public health emergency system can respond and deal with it quickly and efficiently to reduce the negative impact of the incident on public health and ensure social stability and security. The historical indicator data of the public health emergency system is a conditional parameter for constructing a vulnerability measurement model for the public health emergency system. However, before constructing the vulnerability measurement model for the public health emergency system, due to the large amount of historical indicator data, it is necessary to determine the weights of the historical indicator data so that the vulnerability measurement model for the public health emergency system has a focus when making predictions. The determination of weights requires analysis by professionals, so a questionnaire is prepared and published in the target primary medical institutions to conduct an indicator survey to obtain the corresponding survey data. Different historical indicator data will be scored on the questionnaire.

[0088] Furthermore, in a preferred embodiment of the present invention, the principal component analysis algorithm is used to perform weight analysis on the data on the public health emergency survey questionnaire after the index survey to obtain a weight ranking table of historical index data, specifically:

[0089] The principal component analysis method is introduced, and based on the principal component analysis method, the data on the public health emergency survey questionnaire after the indicator survey is converted into a feature matrix, which is calibrated as a type of feature matrix;

[0090] Performing eigenvalue decomposition on the one type of characteristic matrix to obtain a type of eigenvector value, and sorting all the one type of eigenvector values ​​by eigenvalue to obtain a sorting table of the one type of eigenvector values;

[0091] Based on the first type of feature vector value sorting table, weights of different historical indicator data are calculated, and according to the weights of the different historical indicator data, a historical indicator data weight sorting table is constructed.

[0092] It should be noted that after obtaining the data from the questionnaire after the indicator survey, the principal component analysis method can be used to determine the data weight. The principal component analysis method can calculate the eigenvalue of the data to calculate the weight. Therefore, the data needs to be converted into a characteristic matrix and an eigenvector value. Different eigenvector values ​​have different eigenvalues, where the eigenvalue of the eigenvector value can be obtained by calculating its variance value. In the historical indicator data weight sorting table, the eigenvector with a larger eigenvalue, that is, the historical indicator data with a larger eigenvalue, has a greater weight.

[0093] Furthermore, in a preferred embodiment of the present invention, the long short-term memory network algorithm is introduced, and a vulnerability measurement model of the public health emergency system is constructed based on the long short-term memory network algorithm and the weight ranking table of historical indicator data, specifically:

[0094] Presetting a public health emergency system vulnerability prediction time period, and constructing a public health emergency system vulnerability prediction timestamp based on the public health emergency system vulnerability prediction time period;

[0095] Combining the vulnerability prediction timestamp of the public health emergency system and the historical indicator data, constructing historical indicator data in a time series format, calibrating it as historical indicator time series data, and updating the historical indicator time series data based on the historical indicator data weight ranking table to obtain historical indicator weight ranking time series data;

[0096] Introducing a long short-term memory network algorithm, constructing a long short-term memory network basic model, and obtaining an input layer, an LSTM layer, and a fully connected layer in the long short-term memory network basic model;

[0097] Based on the weighted sorting time series data of the historical indicators, the input dimension data of the input layer is determined and calibrated as the target input dimension data;

[0098] Based on the historical data network, when retrieving and building the long short-term memory network model, the standard number of LSTM layers and the standard number of neurons in a single LSTM layer corresponding to different data volumes;

[0099] The data volume in the weighted time series data of the historical indicators is obtained and calibrated as the target data volume. Based on the target data volume, the target number of LSTM layers and the target number of neurons in a single LSTM layer are obtained by combining the standard number of LSTM layers and the standard number of neurons in a single LSTM layer corresponding to different data volumes.

[0100] Based on the target input dimension data, the target number of LSTM layers, and the target number of neurons in a single LSTM layer, the target input layer and the target LSTM layer are constructed. The target input layer, the target LSTM layer, and the single fully connected layer are combined to construct a long short-term memory network training model. The historical indicator weight-sorted time series data is imported into the long short-term memory network training model to obtain a vulnerability measurement model for the public health emergency system.

[0101] It should be noted that the construction of the vulnerability measurement model of the public health emergency system requires the use of the long short-term memory network algorithm, which is a special recursive neural network used to process and predict time series data, and can realize the estimation of the current vulnerability level in the public health emergency system. Since the data processed by the long short-term memory network algorithm is time series data, it is necessary to first determine the predicted time and construct the vulnerability prediction timestamp of the public health emergency system. Through the vulnerability prediction timestamp of the public health emergency system, combined with the historical indicator data, the historical indicator data in the time series format can be obtained. At the same time, since the weight of the historical indicator data is known, it is necessary to combine the weight sorting table of the historical indicator data to obtain the historical indicator weight sorting time series data. The long short-term memory network basic model is a basic model without any training data, in which there are input layers, LSTM layers and fully connected layers. Different data are trained to obtain different prediction results, and the parameters of the input layer, LSTM layer and fully connected layer required are different, that is, the architecture of the long short-term memory network model is different. The parameters of the input layer need to be determined based on time series data, that is, the time series data of historical indicator weight sorting; the parameters of the LSTM layer can refer to the amount of data in the time series data of historical indicator weight sorting. If the amount of data is large, more complex problems can be handled. The number of layers and the number of neurons in a single layer are different for different amounts of data. The fully connected layer is usually single. Therefore, according to the parameters of the input layer, the number of layers of the LSTM layer and the number of neurons in a single layer, and the fully connected layer of a single layer, a long short-term memory network model with a network architecture that meets the requirements can be constructed, and the time series data of historical indicator weight sorting can be imported into the model for data training to obtain a vulnerability measurement model for the public health emergency system.

[0102] In addition, the public health emergency system management method based on model feedback also includes the following steps:

[0103] Obtain public health event cases that have occurred in target primary medical institutions in the big data network and mark them as target public health event cases;

[0104] Analyze the target public health event case to obtain the time period of the target public health event case, wherein the time period of the target public health event case includes three time periods: a pre-event time period, an in-event time period, and a post-event time period, and simultaneously obtain the factor category of the target public health event case, wherein the factor category of the target public health event case includes host factors, environmental factors, and pathogenic factors;

[0105] Based on the time period and factor categories of the target public health event cases, the Haddon matrix is ​​constructed, and the vulnerability analysis of the Haddon matrix in different time periods is carried out through the vulnerability measurement model of the public health emergency system to obtain the vulnerability of different vulnerability assessment areas in the target public health event cases. Among them, the vulnerability of different vulnerability assessment areas in the target public health event cases is the vulnerability of different time periods and different factor categories.

[0106] Analyze the vulnerabilities of the different vulnerability assessment areas in the target public health event cases, and preset vulnerability thresholds of the different vulnerability assessment areas in the target public health event cases, which are calibrated as standard vulnerability thresholds;

[0107] Compare and analyze the vulnerabilities of different vulnerability assessment areas in the target public health event cases with the corresponding standard vulnerability thresholds, and obtain the weak areas of primary medical institutions in the public health emergency system. The weak areas of primary medical institutions in the public health emergency system are the vulnerability assessment areas whose vulnerabilities are not within the standard vulnerability thresholds.

[0108] In the weak areas of primary medical institutions in the public health emergency system, the corresponding time periods and factor categories are marked as weak time periods and weak factor categories, and the intervention methods of the weak time periods and weak factor categories are retrieved and output in the big data network, so that the vulnerability of all vulnerability assessment areas in the target public health event cases is within the standard vulnerability threshold.

[0109] It should be noted that many public health incidents may have occurred in primary medical institutions. It is necessary to select one of the public health incident cases for analysis, so as to analyze the weak areas in primary medical institutions and intervene. The analysis is conducted in three time periods: the pre-incident period, the in-incident period, and the post-incident period, as well as three factor categories: host factors, environmental factors, and pathogenic factors. The pre-incident period is the time period before the accident, involving preventive measures and risk factors. The in-incident period is the time period when the accident occurs, involving protective measures for the incident, etc. The post-incident period is the time period after the accident, involving emergency response and treatment of the incident. Taking the occurrence of new crown infection in primary medical institutions as an example, the constructed Haddon matrix is:

[0110]

[0111]

[0112] After constructing the Haddon matrix, it is necessary to conduct vulnerability analysis on all elements in the Haddon matrix. Vulnerability analysis can be performed using the vulnerability measurement model of the public health emergency system. Since the target primary medical institutions contain different vulnerability assessment areas, and the vulnerabilities corresponding to different time periods and different factor categories in different vulnerability assessment areas are also different, the vulnerability threshold is preset, and the vulnerability in different vulnerability assessment areas is analyzed. Based on the analysis results, the weak areas of primary medical institutions in the public health emergency system can be obtained, and the intervention methods corresponding to the weak time periods and weak factor categories of the weak areas can be retrieved and output in the big data network to achieve the overall ability of the target primary medical institutions to respond to public health emergencies.

[0113] like Figure 3 As shown, the second aspect of the present invention further provides a public health emergency system management system based on model feedback, the public health emergency system management system includes a memory 31 and a processor 32, the memory 31 stores a public health emergency system management method, and when the public health emergency system management method is executed by the processor 32, the following steps are implemented:

[0114] Collect historical indicator data of the public health emergency system, and build a vulnerability measurement model for the public health emergency system based on the indicator data of the public health emergency system;

[0115] Based on the vulnerability measurement model of the public health emergency system, obtain the functional score of the current indicator data of the target public health emergency system, and conduct a vulnerability assessment of the target public health emergency system based on the functional score of the current indicator data;

[0116] Formulate intervention strategies for vulnerable abnormal areas, and output prioritized interventions for vulnerable abnormal areas based on the ranking table of total functional scores of vulnerable abnormal areas.

[0117] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A public health emergency system management method based on model feedback, characterized in that: The following steps are involved: Collect historical indicator data of the public health emergency system, and build a vulnerability measurement model for the public health emergency system based on the indicator data of the public health emergency system; Based on the vulnerability measurement model of the public health emergency system, obtain the functional score of the current indicator data of the target public health emergency system, and conduct a vulnerability assessment of the target public health emergency system based on the functional score of the current indicator data; Formulate intervention strategies for abnormal vulnerability areas, and output intervention strategies for abnormal vulnerability areas in order based on the ranking table of total functional scores of abnormal vulnerability areas; The historical indicator data of the public health emergency system are collected, and a vulnerability measurement model of the public health emergency system is constructed based on the indicator data of the public health emergency system, specifically: The primary medical institutions that need to be managed by the public health emergency system are identified as target primary medical institutions, and the current public health emergency system in the target primary medical institutions is identified as the target public health emergency system; Acquire a historical data network, and acquire historical indicator data of a target public health emergency system through the historical data network, wherein the historical indicator data includes public health data, health infrastructure data, emergency management data of a target primary medical institution, and population and socioeconomic data of a location of the target primary medical institution; Based on the historical indicator data, construct a public health emergency survey questionnaire, and publish the public health emergency survey questionnaire to target primary medical institutions for indicator survey; The principal component analysis method is introduced, and based on the principal component analysis method, the data on the public health emergency survey questionnaire after the indicator survey is converted into a feature matrix, which is calibrated as a type of feature matrix; Performing eigenvalue decomposition on the first-class characteristic matrix to obtain a first-class characteristic vector value, and sorting all the first-class characteristic vector values ​​by eigenvalue to obtain a first-class characteristic vector value sorting table; Based on the first type of feature vector value sorting table, weights of different historical indicator data are calculated, and according to the weights of the different historical indicator data, a historical indicator data weight sorting table is constructed; Introducing the long short-term memory network algorithm, and building a vulnerability measurement model for the public health emergency system based on the long short-term memory network algorithm and the weight ranking table of historical indicator data; Among them, the introduction of the long short-term memory network algorithm, based on the long short-term memory network algorithm and the weight ranking table of historical indicator data, constructs a vulnerability measurement model for the public health emergency system, specifically: Presetting a public health emergency system vulnerability prediction time period, and constructing a public health emergency system vulnerability prediction timestamp based on the public health emergency system vulnerability prediction time period; Combining the vulnerability prediction timestamp of the public health emergency system and the historical indicator data, constructing historical indicator data in a time series format, calibrating it as historical indicator time series data, and updating the historical indicator time series data based on the historical indicator data weight ranking table to obtain historical indicator weight ranking time series data; Introducing a long short-term memory network algorithm, constructing a long short-term memory network basic model, and obtaining an input layer, an LSTM layer, and a fully connected layer in the long short-term memory network basic model; Based on the weighted sorting time series data of the historical indicators, the input dimension data of the input layer is determined and calibrated as the target input dimension data; Based on the historical data network, when retrieving and building the long short-term memory network model, the standard number of LSTM layers and the standard number of neurons in a single LSTM layer corresponding to different data volumes; The data volume in the weighted time series data of the historical indicators is obtained and calibrated as the target data volume. Based on the target data volume, the target number of LSTM layers and the target number of neurons in a single LSTM layer are obtained by combining the standard number of LSTM layers and the standard number of neurons in a single LSTM layer corresponding to different data volumes. Based on the target input dimension data, the target number of LSTM layers, and the target number of neurons in a single LSTM layer, the target input layer and the target LSTM layer are constructed. The target input layer, the target LSTM layer, and the single fully connected layer are combined to construct a long short-term memory network training model. The historical indicator weight-sorted time series data is imported into the long short-term memory network training model to obtain a vulnerability measurement model for the public health emergency system.

2. A public health emergency system management method based on model feedback according to claim 1, characterized in that: The vulnerability measurement model based on the public health emergency system obtains the functional score of the current indicator data of the target public health emergency system, and performs a vulnerability assessment on the target public health emergency system based on the functional score of the current indicator data, specifically: In the target primary medical institution, the current status indicator data of the target public health emergency system is obtained, the vulnerability measurement model of the public health emergency system is run, and the current status indicator data of the target public health emergency system is imported into the vulnerability measurement model of the public health emergency system for storage; In the vulnerability measurement model of the public health emergency response system, a function score qualification threshold is preset, and the function score qualification threshold includes an exposure score qualification threshold, a sensitivity score qualification threshold, and a fitness score qualification threshold; In the vulnerability measurement model of the public health emergency system, a vulnerability assessment is performed on the stored status indicator data of the target public health emergency system to obtain a functional score of the status indicator data of the target public health emergency system, wherein the functional score includes an exposure score, a sensitivity score, and a fitness score; If the function scores of the current status indicator data of the target public health emergency system are all within the qualified function score threshold, the target public health emergency system will be marked as a qualified public health emergency system; If any score in the functional score of the current status indicator data of the target public health emergency system is not within the functional score qualification threshold, the target public health emergency system will be marked as a public health emergency system with abnormal current status, and the public health emergency system with abnormal current status will be divided into regions and the vulnerability assessment of different regions will be ranked.

3. According to the public health emergency system management method based on model feedback described in claim 2, it is characterized in that: The regional division of the abnormal public health emergency response system and the vulnerability assessment ranking of different regions are specifically as follows: Based on the current abnormal public health emergency response system and the public health emergency functionality, the target primary medical institutions are divided into different areas and marked as vulnerability areas to be assessed; Obtaining a functional score of the status indicator data of the current abnormal public health emergency response system, and based on the functional score of the status indicator data of the current abnormal public health emergency response system, marking the vulnerability to be assessed area whose functional score of the status indicator data is maintained within the functional score qualified threshold as a vulnerability qualified area in the vulnerability to be assessed area; In the vulnerability assessment area, the vulnerability assessment area where the function score of the status indicator data does not remain within the qualified function score threshold is marked as an abnormal vulnerability area; The functional scores of different status indicator data in the vulnerability anomaly area are added together to obtain the total functional score of the status indicator data. The total functional scores of the status indicator data of all vulnerability anomaly areas are compared, and a ranking table of the total functional scores of vulnerability anomaly areas is constructed based on the comparison results.

4. According to the public health emergency system management method based on model feedback described in claim 1, it is characterized in that: The intervention strategy for the abnormal vulnerability area is formulated, and based on the ranking table of the total functional scores of the abnormal vulnerability areas, the abnormal vulnerability areas are intervened in sequence, specifically: The functional scores of the status quo indicator data in the vulnerability anomaly area are analyzed, and based on the analysis results, the vulnerability anomaly area is divided into exposure anomaly area, sensitivity anomaly area and fitness anomaly area; For the abnormal exposure area, obtain the current indicator data of the abnormal exposure area, and retrieve the intervention strategy of the abnormal exposure area based on the current indicator data of the abnormal exposure area in the historical data network; For the abnormal sensitivity area, obtain the current status indicator data of the abnormal sensitivity area, and retrieve the intervention strategy of the abnormal sensitivity area based on the current status indicator data of the abnormal sensitivity area in the historical data network; For the fitness abnormal area, obtain the current indicator data of the fitness abnormal area, and retrieve the intervention strategy of the fitness abnormal area based on the current indicator data of the fitness abnormal area in the historical data network; The intervention strategies for the exposure abnormality area, the sensitivity abnormality area and the fitness abnormality area are integrated to obtain the total intervention strategy for the vulnerability abnormality area, and the total intervention strategy for the vulnerability abnormality area is output to different vulnerability abnormality areas in sequence based on the ascending principle in the vulnerability abnormality area function total score ranking table.

5. A public health emergency system management system based on model feedback, characterized in that: The public health emergency system management system includes a memory and a processor, and the memory stores a public health emergency system management method program. When the public health emergency system management method program is executed by the processor, the public health emergency system management method steps as described in any one of claims 1-4 are implemented.

Images