Method for assessing epidemic situation of animal monkeypox outbreak based on monkeypox virus detection data

Abstract

The application discloses a kind of animal monkeypox epidemic situation popular trend assessment method based on monkeypox virus detection data, comprising: obtaining multi-source monkeypox virus detection data and carrying out data preprocessing, determine monkeypox epidemic point, extract the time characteristics and spatial characteristics of each monkeypox epidemic point Cluster monkeypox epidemic points, according to the clustering results in the same cluster Monkeypox epidemic points using correlation analysis to extract the influence factors related to animal monkeypox epidemic;Using graph neural network to analyze the influence factors in space-time, according to the risk space-time characteristics of animal monkeypox epidemic, analyze and define the corresponding space-time factor;According to the influence factors and space-time factors, a monkeypox risk prediction model is constructed, the risk prediction level of the animal monkeypox epidemic of the target monkeypox epidemic point is obtained to assess the popular trend.The present application is based on the space-time factor and influence factor of monkeypox epidemic, and the popular trend is predicted and evaluated, which provides a scientific basis for the rational allocation of medical and health resources.

Description

Technical Field

[0001] The present invention relates to the field of deep learning and epidemiological technology, and more specifically, to a method for assessing the epidemic situation of animal monkeypox based on monkeypox virus detection data. Background Art

[0002] Monkeypox is a rare zoonotic disease characterized by a skin rash. It is caused by the monkeypox virus, a member of the genus Orthopoxvirus in the family Poxviridae. Monkeypox virus can be transmitted from person to person and from animal to person. Genomic sequencing analysis of viral DNA has shown that monkeypox virus consists of two main subtypes (monkeypox virus type 1 and monkeypox virus type 2), with monkeypox virus type 1 currently prevalent. WHO is currently developing or partially developing interim guidelines for the prevention and control of monkeypox virus to support Member States in raising awareness, surveillance, laboratory diagnosis and testing, case investigation and contact tracing, clinical management and infection prevention and control, vaccines and immunization, and risk communication and community engagement.

[0003] Traditional infectious disease forecasting relies primarily on traditional epidemiological models and statistical methods, such as SEIR and ARIMA models. However, these methods rely on assumptions and simplifications, failing to fully capture the complex factors and nonlinear relationships in epidemic spread. In recent years, with the development of artificial intelligence (AI), deep learning models have demonstrated strong potential in epidemic forecasting. Deep learning models can automatically learn and extract high-level features from epidemic data, enabling a better understanding and prediction of epidemic patterns and trends. Monkeypox epidemic data can be subject to interference and noise from various factors, such as reporting delays, missing data, and erroneous reporting. The stability and robustness of the model also require consideration. Therefore, aggregating multi-source monkeypox virus detection data to assess the epidemic situation is an urgent issue that needs to be addressed. Summary of the Invention

[0004] In order to solve the above technical problems, the present invention proposes a method for evaluating the epidemic situation of animal monkeypox epidemics based on monkeypox virus detection data. Based on the risk spatiotemporal transmission characteristics of the animal monkeypox epidemic, the spatiotemporal correlation of the animal monkeypox epidemic is analyzed and the corresponding spatiotemporal factors and influencing factors are defined, and the epidemic situation is predicted and evaluated, providing a scientific basis for the rational allocation and utilization of medical and health resources.

[0005] The present invention provides a method for assessing the epidemic situation of monkeypox in animals based on monkeypox virus detection data, comprising the following steps:

[0006] Acquire multi-source monkeypox virus detection data and perform data preprocessing, determine monkeypox epidemic sites based on the preprocessed multi-source monkeypox virus detection data, and extract the temporal and spatial characteristics of each monkeypox epidemic site;

[0007] Clustering the monkeypox epidemic sites based on the temporal and spatial characteristics, and extracting influencing factors related to the animal monkeypox epidemic using correlation analysis for the monkeypox epidemic sites in the same cluster according to the clustering results;

[0008] Using graph neural networks to conduct spatiotemporal analysis of the influencing factors, we analyzed the spatiotemporal correlation of the monkeypox epidemic and defined corresponding spatiotemporal factors based on the spatiotemporal risk characteristics of the epidemic.

[0009] A monkeypox risk prediction model is constructed based on the influencing factors and spatiotemporal factors to obtain the risk prediction level of the animal monkeypox epidemic at the target monkeypox epidemic site, and the epidemic situation is evaluated based on the changes between the risk prediction level and the current risk level.

[0010] In this scheme, monkeypox epidemic sites are determined based on pre-processed multi-source monkeypox virus detection data, and the temporal and spatial characteristics of each monkeypox epidemic site are extracted, specifically:

[0011] Obtaining monkeypox virus detection data aggregated by different detection methods within a preset time period, standardizing and dimensionality reduction processing the collected multi-source monkeypox virus detection data, and screening positive samples from the pre-processed multi-source monkeypox virus detection data;

[0012] Extracting geographic location information from the screened positive samples, integrating the geographic location information after traversing all positive samples, using the integrated geographic location information to divide the positive samples into positive sample subsets, and determining the monkeypox epidemic point based on the geographic location labels of each positive sample subset;

[0013] Obtaining a time series sequence of positive samples from each monkeypox epidemic site, extracting trend turning points from the positive sample time series sequence, dynamically segmenting the positive sample time series sequence using the trend turning points, extracting trend feature identifiers from each segmented subsegment, characterizing the temporal clustering of positive samples using the trend feature identifiers, and generating temporal features for each monkeypox epidemic site;

[0014] Obtaining a global spatial distribution based on the geographic location information of each monkeypox epidemic site, calculating a spatial weight for each monkeypox epidemic site using spatial distance, characterizing the spatial relationship between the monkeypox epidemic sites using the spatial weight, and screening neighboring epidemic sites based on the spatial relationship;

[0015] Based on each monkeypox epidemic point and its corresponding neighboring epidemic point, the local spatial distribution is obtained. The global spatial distribution and the local spatial distribution are introduced into the CNN network to mine the spatial clustering of positive samples and generate the spatial characteristics of each monkeypox epidemic point.

[0016] In this scheme, monkeypox epidemic sites are clustered based on the temporal and spatial characteristics, specifically:

[0017] Obtain the temporal and spatial characteristics of each monkeypox epidemic site at a preset time step, sort the monkeypox epidemic sites according to the number of positive samples, and select a preset number of monkeypox epidemic sites from the sorting results as initial cluster centers;

[0018] A metric function was constructed using Mahalanobis distance. The Mahalanobis distance was calculated based on the temporal and spatial characteristics of the monkeypox epidemic sites to characterize the similarity between the sites. Other monkeypox epidemic sites were assigned to the initial cluster center with the closest distance to generate the corresponding clusters.

[0019] The cluster centers and clustering results are continuously updated through iterative calculations. After obtaining a new cluster center in each iteration, the distance between the monkeypox epidemic site and its cluster center is compared with the distance from the previous monkeypox epidemic site to the cluster center.

[0020] If it is less than or equal to, the monkeypox epidemic point will be retained in the cluster center to which it belongs. Otherwise, the distance will be recalculated and the monkeypox epidemic point will be assigned to the cluster center with the nearest distance. When the iteration number threshold is met, the clustering result of the last iteration will be selected as the clustering result of the monkeypox epidemic point.

[0021] In this scheme, based on the clustering results, correlation analysis is used to extract the influencing factors related to the monkeypox epidemic in animals from the monkeypox epidemic sites in the same cluster, specifically:

[0022] According to the clustering results of monkeypox epidemic points, the monkeypox epidemic point clusters with spatiotemporal aggregation within the preset time step are obtained. Monkeypox epidemic points in the same cluster indicate the existence of spatiotemporal aggregation;

[0023] Using big data and knowledge graphs to obtain relevant datasets of animal monkeypox outbreaks, using maximum relevance and minimum redundancy to perform feature selection in the relevant datasets, and generating preliminary influencing factors for the occurrence of animal monkeypox outbreaks at monkeypox outbreak sites belonging to the same cluster;

[0024] Obtaining a set of preliminary influencing factors corresponding to all clusters in the clustering results, collecting parameter samples from the corresponding monkeypox epidemic sites based on the preliminary influencing factor set, generating a parameter sample data set for each cluster, performing data balancing processing, and generating participants;

[0025] The federation idea is introduced into the particle swarm algorithm to improve it. The particle swarm algorithm is initialized, and the parameter sample data set of each participant is used as the algorithm input. The particle population is set and the particle position is decoded. The particle fitness is calculated based on the classification accuracy of the combination of influencing factors.

[0026] The position is updated based on the comparison of particle fitness, and the best combination of influencing factors of each participant is obtained through iterative update, and the best combination of influencing factors of each participant is transmitted to the preset aggregation server;

[0027] After receiving the influencing factor combinations of all participants, the preset aggregation server shares the influencing factor combinations with all participants, re-updates the particle swarm, and performs iterative updates until the termination condition is met, and finally obtains the optimal influencing factor combination of each participant.

[0028] In this solution, a graph neural network is used to perform spatiotemporal analysis of the influencing factors, specifically:

[0029] An undirected graph was constructed based on the spatial distribution of monkeypox epidemic sites. Each monkeypox epidemic site was used as a node to obtain the optimal combination of influencing factors corresponding to the monkeypox epidemic site. Parameter samples were obtained based on the optimal combination of influencing factors to generate additional features for the nodes. The edge structure between the nodes was set based on whether there was spatiotemporal clustering between the monkeypox epidemic sites.

[0030] Neighborhood nodes are set according to the monkeypox epidemic points with spatiotemporal clustering, an adjacency matrix is ​​obtained to represent the undirected graph, and a graph neural network is used for learning representation. During the representation process, attention weights are introduced to weight the neighborhood nodes of the monkeypox epidemic points. The node representation is updated using spatiotemporal convolution and neighbor aggregation to extract the spatiotemporal characteristics of the influencing factors corresponding to the monkeypox epidemic points.

[0031] In this plan, based on the spatiotemporal characteristics of the risk of monkeypox epidemics in animals, the spatiotemporal correlation of the epidemic in animals is analyzed and the corresponding spatiotemporal factors are defined, specifically:

[0032] Match the spatiotemporal characteristics of various influencing factors of the monkeypox epidemic site with the temporal characteristics and spatial characteristics of the time series of positive samples of the monkeypox epidemic site to obtain the temporal correlation and spatial correlation respectively;

[0033] Normalization is performed on the time correlation and the space correlation to obtain corresponding time factors and space factors, and the time factors and the space factors are added together to obtain the spatiotemporal factors of each influencing factor.

[0034] In this scheme, a monkeypox risk prediction model is constructed based on the influencing factors and spatiotemporal factors to obtain the risk prediction level of the animal monkeypox epidemic at the target monkeypox epidemic site, specifically:

[0035] A monkeypox risk prediction model was constructed using a multi-layer perceptron based on the influencing factors of monkeypox epidemic sites and the corresponding spatiotemporal factors. Training samples were constructed based on the influencing factor parameters with monkeypox epidemic risk level labels at monkeypox epidemic sites. The monkeypox risk prediction model was then trained and tested.

[0036] The spatiotemporal characteristics of the parameter samples corresponding to the current influencing factors of the target monkeypox epidemic site are obtained, and the spatiotemporal factors are combined with the spatiotemporal factors to be imported into the monkeypox risk prediction model, and the fully connected layer is used to output the risk prediction level of the animal monkeypox epidemic after a preset time.

[0037] In this solution, the risk deviation between the predicted risk level of the target monkeypox epidemic site and the current risk level is obtained, and the ratio of the risk deviation to the preset time period is calculated as the risk change rate;

[0038] Based on the current risk level, a risk change benchmark value is obtained according to historical animal monkeypox epidemic examples, the risk change rate is compared with the risk change benchmark value, and the epidemic situation is evaluated based on the comparison deviation combined with the preset threshold interval difference.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] Based on the spatiotemporal transmission characteristics of the risk of animal monkeypox epidemics, the present invention analyzes the spatiotemporal correlation of animal monkeypox epidemics and defines the corresponding spatiotemporal factors and influencing factors. It constructs a monkeypox risk prediction model to accurately predict the risk level of monkeypox epidemic points, and predicts and evaluates the epidemic situation, providing a scientific basis for the rational allocation and utilization of medical and health resources and the implementation of monkeypox prevention and control work. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments or exemplary embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments or exemplary descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained according to these drawings without paying any creative work.

[0042] Figure 1 A flow chart showing a method for assessing the epidemic situation of animal monkeypox based on monkeypox virus detection data is shown;

[0043] Figure 2 A schematic diagram of the process of extracting influencing factors related to the animal monkeypox epidemic is shown;

[0044] Figure 3 A schematic diagram of the process of using graph neural networks to perform spatiotemporal analysis of influencing factors is shown;

[0045] Figure 4 A block diagram of an animal monkeypox epidemic situation assessment system based on monkeypox virus detection data is shown. DETAILED DESCRIPTION

[0046] In order to more clearly understand the above-mentioned objects, 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, in the absence of conflict, the embodiments of the present application and the features therein can be combined with each other.

[0047] 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 scope of protection of the present invention is not limited to the specific embodiments disclosed below.

[0048] like Figure 1 As shown, an embodiment of the present invention provides a method for assessing the epidemic situation of animal monkeypox based on monkeypox virus detection data, comprising:

[0049] S102, obtaining multi-source monkeypox virus detection data and performing data preprocessing, determining monkeypox epidemic sites based on the preprocessed multi-source monkeypox virus detection data, and extracting temporal and spatial characteristics of each monkeypox epidemic site;

[0050] S104, clustering the monkeypox epidemic sites based on the temporal and spatial characteristics, and extracting influencing factors related to the animal monkeypox epidemic using correlation analysis for the monkeypox epidemic sites in the same cluster according to the clustering results;

[0051] S106, using a graph neural network to perform spatiotemporal analysis on the influencing factors, analyzing the spatiotemporal correlation of the animal monkeypox epidemic according to the spatiotemporal risk characteristics of the animal monkeypox epidemic and defining corresponding spatiotemporal factors;

[0052] S108, constructing a monkeypox risk prediction model based on the influencing factors and spatiotemporal factors, obtaining the risk prediction level of the animal monkeypox epidemic at the target monkeypox epidemic site, and evaluating the epidemic situation based on the change between the risk prediction level and the current risk level.

[0053] It should be noted that common monkeypox detection methods typically include nucleic acid testing, laboratory testing, clinical observation, and imaging. Diagnosis of monkeypox primarily relies on a combination of clinical symptoms and laboratory testing, with PCR being the most commonly used and accurate method. Early detection and timely treatment are crucial for disease control.

[0054] Obtaining monkeypox virus detection data aggregated by different detection methods within a preset time period, standardizing and dimensionality reduction processing the collected multi-source monkeypox virus detection data, and screening and determining positive samples from the pre-processed multi-source monkeypox virus detection data; extracting geographic location information from the screened positive samples, integrating the geographic location information after traversing all positive samples, using the integrated geographic location information to divide the positive samples into positive sample subsets, and determining the monkeypox epidemic point based on the geographic location label of each positive sample subset;

[0055] Obtain a time series sequence of positive samples from each monkeypox epidemic site, and extract trend turning points from the positive sample time series sequence. The trend point is the turning point where the time series transforms from one form to another, and is the change point of the local trend. Use the trend turning points to dynamically segment the positive sample time series sequence, and extract trend feature identifiers of each segment after segmentation. The trend feature identifiers include trend distance and trend morphology. The trend distance is the difference between the start and end points of the sequence segment, which characterizes the trend direction and amplitude of the sequence segment. The trend morphology characterizes sequence segments with the same (or similar) trends but different internal morphologies. Its essence is to measure the sequence segments through their trend points. The trend feature identifier is used to characterize the temporal clustering of positive samples, and the number of cases corresponding to the case sample sequence is read to determine whether it has a trend of increasing, decreasing, or remaining unchanged over time, thereby generating the temporal characteristics of each monkeypox epidemic point; the global spatial distribution is obtained based on the geographical location information of each monkeypox epidemic point, the spatial weight of each monkeypox epidemic point is calculated using the spatial distance, the spatial weight is used to characterize the spatial relationship between the monkeypox epidemic points, and the neighboring epidemic points are screened based on the spatial relationship; the local spatial distribution is obtained based on each monkeypox epidemic point and its corresponding neighboring epidemic point, the global spatial distribution and the local spatial distribution are introduced into a CNN network to mine the spatial clustering of positive samples, and the spatial characteristics of each monkeypox epidemic point are generated.

[0056] It should be noted that the time characteristics and spatial characteristics of each monkeypox epidemic point at a preset time step are obtained, the monkeypox epidemic points are sorted according to the number of positive samples, and a preset number of monkeypox epidemic points are selected from the sorting results as initial cluster centers; a metric function is constructed through the Mahalanobis distance, and the Mahalanobis distance is calculated according to the time characteristics and spatial characteristics between the monkeypox epidemic points to characterize the similarity between the monkeypox epidemic points, and the other monkeypox epidemic points are attributed to the initial cluster center with the nearest distance to generate corresponding clusters; the cluster centers and clustering results are continuously updated through iterative calculations, and after obtaining a new cluster center in each iteration, the distance between the monkeypox epidemic point and its cluster center is compared with the distance from the previous monkeypox epidemic point to the cluster center; if it is less than or equal to, the monkeypox epidemic point is retained in the cluster center to which it belongs, otherwise the distance is recalculated and the monkeypox epidemic point is attributed to the cluster center with the nearest distance. When the iteration number threshold is met, the clustering result of the last iteration is selected as the clustering result of the monkeypox epidemic point.

[0057] Figure 2 A schematic diagram of the process for extracting influencing factors related to animal monkeypox epidemics is shown.

[0058] According to an embodiment of the present invention, the influencing factors related to the monkeypox epidemic in animals are extracted by using correlation analysis for the monkeypox epidemic sites in the same cluster based on the clustering results, specifically:

[0059] S202, obtaining, based on the clustering results of the monkeypox epidemic points, a cluster of monkeypox epidemic points that is spatiotemporally clustered within a preset time step, wherein the monkeypox epidemic points in the same cluster indicate spatiotemporal clustering;

[0060] S204, using big data and knowledge graphs to obtain a dataset related to the occurrence of monkeypox in animals, performing feature selection on the dataset using maximum relevance and minimum redundancy, and generating preliminary influencing factors for the occurrence of monkeypox in animals at the monkeypox epidemic sites in the same cluster;

[0061] S206, obtaining a set of preliminary influencing factors corresponding to all clusters in the clustering results, collecting parameter samples from the corresponding monkeypox epidemic sites based on the set of preliminary influencing factors, generating a parameter sample data set for each cluster, performing data balancing processing, and generating participating parties;

[0062] S208, introducing the federation concept into the particle swarm algorithm to improve it, initializing the particle swarm algorithm, taking the parameter sample data set of each participant as the algorithm input, setting the particle population and decoding the particle position, and calculating the particle fitness based on the classification accuracy of the influencing factor combination;

[0063] S210, updating the positions according to the comparison of the particle fitness, obtaining the best combination of influencing factors of each participant through iterative updating, and transmitting the best combination of influencing factors of each participant to the preset aggregation server;

[0064] S212, after receiving the influencing factor combinations of all participants, the preset aggregation server shares the influencing factor combinations with all participants, re-updates the particle swarm, and performs iterative updates until the termination condition is met, and finally obtains the optimal influencing factor combination of each participant.

[0065] It's important to note that big data and knowledge graphs indicate that factors influencing monkeypox outbreaks in animals include infection routes, viral load, individual immunity, and environmental factors. Because different clusters have different influencing factors, applying standard feature selection algorithms to imbalanced data often results in features that prioritize the majority class over the minority class. Therefore, data balancing is performed to improve the applicability and comprehensiveness of factor extraction. In the federated particle swarm algorithm, the commonly used KNN classifier is selected, and the classification accuracy of each particle is calculated as its fitness. The particle swarm algorithm parameters are initialized, including the size of the particle swarm, maximum particle velocity, maximum number of iterations, and learning factor. A swarm is randomly generated, and the fitness of each particle is calculated according to the particle fitness function. The optimal fitness of each particle is updated, and the current optimal position of each particle is used as the initial position to find the global optimal position. The particle's velocity and position are updated using the particle itself and the global optimal value, and the fitness of the updated particle is calculated. During the iteration process, the optimal influencing factor combination of each participant and its classification accuracy are transmitted to the aggregation server. The aggregation server then transmits M-1 optimal influencing factor combinations to each participant, where M is the total number of participants. The multi-party federated evolutionary feature selection based on particle swarm optimization improves the stability and robustness of the subsequent model and enhances the predictive performance of the influencing factors selected by each participant.

[0066] Figure 3 A schematic diagram of the process of using graph neural networks to perform spatiotemporal analysis of influencing factors is shown.

[0067] According to an embodiment of the present invention, a graph neural network is used to perform spatiotemporal analysis on the influencing factors, specifically:

[0068] S302, constructing an undirected graph based on the spatial distribution of monkeypox epidemic sites, using each monkeypox epidemic site as a node, obtaining the optimal influencing factor combination corresponding to the monkeypox epidemic site, obtaining parameter samples based on the optimal influencing factor combination to generate additional features for the node, and setting the edge structure between the nodes based on whether there is spatiotemporal clustering between the monkeypox epidemic sites;

[0069] S304, setting neighborhood nodes based on the monkeypox epidemic points with spatiotemporal clustering, obtaining an adjacency matrix to represent the undirected graph, using a graph neural network for learning representation, introducing attention weights in the representation process to weight the neighborhood nodes of the monkeypox epidemic points, using spatiotemporal convolution and neighbor aggregation to update the node representation, and extracting the spatiotemporal characteristics of the influencing factors corresponding to the monkeypox epidemic points.

[0070] It should be noted that since the aggregation of monkeypox in the time dimension, space dimension and time-space dimension is usually caused by the time correlation and space correlation of the epidemic, there is a certain time-space correlation in the spread and diffusion of the monkeypox epidemic. The spatiotemporal characteristics of each influencing factor of the monkeypox epidemic point are matched with the time characteristics and spatial characteristics of the time series of positive samples of the monkeypox epidemic point to obtain the time correlation and spatial correlation respectively; determine whether the change of the influencing factor is consistent with the development and change of the number of positive samples or the epidemic risk level. If the time series change of the influencing factor within the preset time is not correlated with the time series change of the number of positive samples, then the time correlation of the influencing factor is 0. The time correlation and spatial correlation are normalized to obtain the corresponding time factor and space factor, and the time factor and space factor are added to obtain the spatiotemporal factor of each influencing factor.

[0071] A monkeypox risk prediction model is constructed using a multi-layer perceptron based on the influencing factors of monkeypox epidemic sites and the spatiotemporal factors corresponding to the influencing factors. The monkeypox risk prediction model incorporates the spatiotemporal factors of the epidemic risk, thereby taking into account the possible impact of the influencing factors and the spatiotemporal autocorrelation of the epidemic on the occurrence of animal monkeypox epidemics. Training samples are constructed based on the influencing factor parameters with monkeypox epidemic risk level labels at monkeypox epidemic sites, and the monkeypox risk prediction model is trained and tested; preferably, risk clustering is performed on historical animal monkeypox epidemics, and three statistically significant results are distinguished by clustering: high-risk level clustering, medium-risk level clustering, and low-risk level clustering. Data samples are constructed based on the indicator parameters of the influencing factor indicators of the risk level labels, and the data samples are matched with the risk level labels to generate training samples for training the multi-layer perceptron. The spatiotemporal characteristics of the parameter samples corresponding to the current influencing factors of the target monkeypox epidemic site are obtained, combined with the spatiotemporal factors and imported into the monkeypox risk prediction model, and the fully connected layer is used to output the risk prediction level of the animal monkeypox epidemic after a preset time.

[0072] Obtain the risk deviation between the predicted risk level of the target monkeypox epidemic site and the current risk level, calculate the ratio of the risk deviation to the preset time period as the risk change rate; based on the current risk level, obtain the historical risk change situation according to historical animal monkeypox epidemic examples, calculate the risk change benchmark value according to the historical risk change situation, compare the risk change rate with the risk change benchmark value, and when the risk change rate is greater than the risk change benchmark value, it proves that a high epidemic situation exists, obtain the preset threshold interval where the comparison deviation falls, and obtain the level of the high epidemic situation according to the level corresponding to the preset threshold interval.

[0073] Figure 4 A block diagram of the animal monkeypox epidemic situation assessment system based on monkeypox virus detection data of the present invention is shown.

[0074] The second aspect of the present invention also provides an animal monkeypox epidemic situation assessment system 4 based on monkeypox virus detection data, which includes: a memory 41 and a processor 42, wherein the memory and the processor store and execute an animal monkeypox epidemic situation assessment method program based on monkeypox virus detection data.

[0075] In the several embodiments provided in this application, it should be understood that the disclosed methods and systems can be implemented in other ways. The system embodiments described above are merely illustrative. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored, or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms. In addition, the functional units in the various embodiments of the present invention can all be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the above-mentioned integrated units can be implemented in the form of hardware or in the form of hardware plus software functional units.

[0076] Those skilled in the art will understand that all or part of the steps of the above-mentioned method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above-mentioned method embodiments. The aforementioned storage medium includes various media that can store program codes, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks or optical disks. Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods described in various embodiments of the present invention. The aforementioned storage medium includes various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0077] The above description is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed by the present invention, which should be covered by the scope of protection of the present invention.

Claims

1. A method for assessing the epidemic situation of animal monkeypox based on monkeypox virus detection data, characterized in that: The following steps are involved: Acquire multi-source monkeypox virus detection data and perform data preprocessing, determine monkeypox epidemic sites based on the preprocessed multi-source monkeypox virus detection data, and extract the temporal and spatial characteristics of each monkeypox epidemic site; Clustering the monkeypox epidemic sites based on the temporal and spatial characteristics, and extracting influencing factors related to the animal monkeypox epidemic using correlation analysis for the monkeypox epidemic sites in the same cluster according to the clustering results; Using graph neural networks to conduct spatiotemporal analysis of the influencing factors, we analyzed the spatiotemporal correlation of the monkeypox epidemic and defined corresponding spatiotemporal factors based on the spatiotemporal risk characteristics of the epidemic. Constructing a monkeypox risk prediction model based on the influencing factors and spatiotemporal factors, obtaining a predicted risk level of the animal monkeypox epidemic at the target monkeypox epidemic site, and evaluating the epidemic situation based on the change between the predicted risk level and the current risk level; Graph neural networks are used to perform spatiotemporal analysis of the influencing factors, specifically: An undirected graph was constructed based on the spatial distribution of monkeypox epidemic sites. Each monkeypox epidemic site was used as a node to obtain the optimal combination of influencing factors corresponding to the monkeypox epidemic site. Parameter samples were obtained based on the optimal combination of influencing factors to generate additional features for the nodes. The edge structure between the nodes was set based on whether there was spatiotemporal clustering between the monkeypox epidemic sites. Neighborhood nodes are set according to the monkeypox epidemic points with spatiotemporal clustering, an adjacency matrix is ​​obtained to represent the undirected graph, a graph neural network is used for learning representation, attention weights are introduced in the representation process to weight the neighborhood nodes of the monkeypox epidemic points, spatiotemporal convolution and neighbor aggregation are used to update the node representation, and the spatiotemporal characteristics of the influencing factors corresponding to the monkeypox epidemic points are extracted; According to the spatiotemporal characteristics of the risk of monkeypox in animals, the spatiotemporal correlation of the epidemic in animals was analyzed and the corresponding spatiotemporal factors were defined, specifically: Match the spatiotemporal characteristics of each influencing factor of the monkeypox epidemic site with the temporal characteristics and spatial characteristics of the time series of the medical record samples of the monkeypox epidemic site to obtain the temporal correlation and spatial correlation respectively; Normalization is performed on the time correlation and the space correlation to obtain corresponding time factors and space factors, and the time factors and the space factors are added together to obtain the spatiotemporal factors of each influencing factor.

2. The method for assessing the epidemic situation of monkeypox in animals based on monkeypox virus detection data according to claim 1, characterized in that: The monkeypox epidemic sites were determined based on the pre-processed multi-source monkeypox virus detection data, and the temporal and spatial characteristics of each monkeypox epidemic site were extracted, specifically: Obtain monkeypox virus detection data aggregated by different detection methods within a preset time period, standardize and perform dimensionality reduction processing on the collected multi-source monkeypox virus detection data, and screen medical record samples from the pre-processed multi-source monkeypox virus detection data; Extracting geographic location information from the screened medical record samples, integrating the geographic location information after traversing all medical record samples, using the integrated geographic location information to divide the medical record samples into medical record sample subsets, and determining the monkeypox epidemic point based on the geographic location label of each medical record sample subset; Obtaining a time series of medical record samples from each monkeypox epidemic site, extracting trend turning points from the time series of medical record samples, dynamically segmenting the time series of medical record samples using the trend turning points, extracting trend feature identifiers for each segmented subsegment, characterizing the temporal clustering of the medical record samples using the trend feature identifiers, and generating temporal features for each monkeypox epidemic site; Obtaining a global spatial distribution based on the geographic location information of each monkeypox epidemic site, calculating a spatial weight for each monkeypox epidemic site using spatial distance, characterizing the spatial relationship between the monkeypox epidemic sites using the spatial weight, and screening neighboring epidemic sites based on the spatial relationship; Based on each monkeypox epidemic point and its corresponding neighboring epidemic point, the local spatial distribution is obtained. The global spatial distribution and the local spatial distribution are introduced into the CNN network to mine the spatial clustering of medical record samples and generate the spatial characteristics of each monkeypox epidemic point.

3. The method for assessing the epidemic situation of monkeypox in animals based on monkeypox virus detection data according to claim 1, characterized in that: The monkeypox epidemic sites were clustered based on the temporal and spatial characteristics, specifically: Obtain the temporal and spatial characteristics of each monkeypox epidemic site at a preset time step, sort the monkeypox epidemic sites according to the number of medical record samples, and select a preset number of monkeypox epidemic sites from the sorting results as initial cluster centers; A metric function was constructed using Mahalanobis distance. The Mahalanobis distance was calculated based on the temporal and spatial characteristics of the monkeypox epidemic sites to characterize the similarity between the sites. Other monkeypox epidemic sites were assigned to the initial cluster center with the closest distance to generate the corresponding clusters. The cluster centers and clustering results are continuously updated through iterative calculations. After obtaining a new cluster center in each iteration, the distance between the monkeypox epidemic site and its cluster center is compared with the distance from the previous monkeypox epidemic site to the cluster center. If it is less than or equal to, the monkeypox epidemic point will be retained in the cluster center to which it belongs. Otherwise, the distance will be recalculated and the monkeypox epidemic point will be assigned to the cluster center with the nearest distance. When the iteration number threshold is met, the clustering result of the last iteration will be selected as the clustering result of the monkeypox epidemic point.

4. The method for assessing the epidemic situation of monkeypox in animals based on monkeypox virus detection data according to claim 1, characterized in that: Based on the clustering results, correlation analysis was used to extract the influencing factors related to the monkeypox epidemic in the same cluster, specifically: According to the clustering results of monkeypox epidemic points, the monkeypox epidemic point clusters with spatiotemporal aggregation within the preset time step are obtained. Monkeypox epidemic points in the same cluster indicate the existence of spatiotemporal aggregation; Using big data and knowledge graphs to obtain relevant datasets of animal monkeypox outbreaks, using maximum relevance and minimum redundancy to perform feature selection in the relevant datasets, and generating preliminary influencing factors for the occurrence of animal monkeypox outbreaks at monkeypox outbreak sites belonging to the same cluster; Obtaining a set of preliminary influencing factors corresponding to all clusters in the clustering results, collecting parameter samples from the corresponding monkeypox epidemic sites based on the preliminary influencing factor set, generating a parameter sample data set for each cluster, performing data balancing processing, and generating participants; The federation idea is introduced into the particle swarm algorithm to improve it. The particle swarm algorithm is initialized, and the parameter sample data set of each participant is used as the algorithm input. The particle population is set and the particle position is decoded. The particle fitness is calculated based on the classification accuracy of the combination of influencing factors. The position is updated based on the comparison of particle fitness, and the best combination of influencing factors of each participant is obtained through iterative update, and the best combination of influencing factors of each participant is transmitted to the preset aggregation server; After receiving the influencing factor combinations of all participants, the preset aggregation server shares the influencing factor combinations with all participants, re-updates the particle swarm, and performs iterative updates until the termination condition is met, and finally obtains the optimal influencing factor combination of each participant.

5. The method for assessing the epidemic situation of monkeypox in animals based on monkeypox virus detection data according to claim 1, characterized in that: A monkeypox risk prediction model was constructed based on the influencing factors and spatiotemporal factors to obtain the risk prediction level of the animal monkeypox epidemic at the target monkeypox epidemic site, specifically: A monkeypox risk prediction model was constructed using a multi-layer perceptron based on the influencing factors of monkeypox epidemic sites and the corresponding spatiotemporal factors. Training samples were constructed based on the influencing factor parameters with monkeypox epidemic risk level labels at monkeypox epidemic sites. The monkeypox risk prediction model was then trained and tested. The spatiotemporal characteristics of the parameter samples corresponding to the current influencing factors of the target monkeypox epidemic site are obtained, and the spatiotemporal factors are combined with the spatiotemporal factors to be imported into the monkeypox risk prediction model, and the fully connected layer is used to output the risk prediction level of the animal monkeypox epidemic after a preset time.

6. The method for assessing the epidemic situation of monkeypox in animals based on monkeypox virus detection data according to claim 5, characterized in that: Obtaining the risk deviation between the predicted risk level and the current risk level of the target monkeypox epidemic site, and calculating the ratio of the risk deviation to a preset time period as the risk change rate; Based on the current risk level, a risk change benchmark value is obtained according to historical animal monkeypox epidemic examples, the risk change rate is compared with the risk change benchmark value, and the epidemic situation is evaluated based on the comparison deviation combined with the preset threshold interval difference.

Images