A method for small-scale group contact risk assessment and infectious disease transmission prediction

Abstract

The purpose of the present invention is to provide a method for assessing the risk of group contact at a small spatial scale and predicting the spread of infectious diseases, which relates to the technical field of infectious disease epidemic prediction. Step 1: Use a wearable device to measure the contact behavior data between individuals in a target place within one day; Step 2: Create an adjacency matrix based on the pre-processed contact behavior data to abstractly represent the contact behavior between each individual; Establish an undirected weighted contact network between groups within a small spatial scale; Step 3: Fusion the ABM model with the undirected weighted contact network to obtain an ABM model based on the contact network; Step 4: By setting the infectious disease transmission parameters and fitting the parameters of the ABM model, the infectious disease outbreak risk assessment result is obtained. The present invention can not only break through the limitations of the traditional age matrix and provide a more accurate prediction of infectious disease transmission, but also provide accurate situation analysis and intervention measures effect evaluation before or in the early stage of an epidemic outbreak.

Description

Technical Field

[0001] The present invention relates to the technical field of infectious disease epidemic prediction, and in particular to a method for small-scale group contact risk assessment and infectious disease transmission prediction. Background Art

[0002] With the acceleration of globalization and the increase in population density, infectious diseases, especially airborne respiratory infectious diseases, have become one of the factors that pose a major threat to public health and social economy. In recent years, multiple large-scale infectious disease outbreaks such as SARS, H1N1 influenza, and COVID-19 have exposed the shortcomings of existing infectious disease monitoring and prevention and control methods. In order to effectively respond to possible infectious disease outbreaks in the future, especially in small spaces such as key places such as schools, hospitals, and nursing homes, it is crucial to conduct more accurate infectious disease transmission prediction and risk assessment. However, existing technologies have many deficiencies in infectious disease transmission prediction and risk assessment, which need to be improved urgently.

[0003] The spread of infectious diseases usually depends on close contact between people. There are mainly the following traditional infectious disease models: 1. Infectious disease spread prediction method based on traditional SEIR model: SEIR model (susceptible-exposed-infected-recovered model) is a classic infectious disease dynamics model. This model predicts the spread trajectory of the disease by dividing the population into different states (susceptible S, exposed E, infected I, recovered R) and simulating the transformation between different states. SEIR model has certain applications in large-scale epidemic prediction, such as analyzing the overall development trend of the epidemic at the national or provincial scale. 2. Model based on uniform mixing assumption: Traditional infectious disease models usually adopt the "uniform mixing" assumption, that is, assuming that the contact between individuals is random and uniform, and each individual has an equal chance of contacting other individuals. This model is more applicable to describing the spread of large-scale, macro-scale epidemics, but in small-space places, such as schools, hospitals, factories and other closed environments, this assumption seems too idealistic and ignores the heterogeneity of contact behavior between individuals, that is, there are significant differences in contact frequency and contact intensity between different individuals. 3. Infectious disease models based on large-scale social contact data: With the development of data collection technology, more and more studies are building more accurate infectious disease transmission models based on social contact data. For example, through questionnaires, contact diaries, etc., social contact behavior data between individuals in the population are collected to build infectious disease transmission models. This type of research has been widely used in large-scale regions such as the country, province, and city, such as modeling the contact behavior of people of different genders and ages in a city and predicting the spread of influenza. However, most of these studies are concentrated in large-scale spatial scopes, while research on small spatial scales and key places such as schools, hospitals, and nursing homes is still relatively lacking. 4. Non-drug interventions (NPIs) combined with models: In the face of infectious disease outbreaks, non-drug interventions (NPIs) such as reducing crowd gatherings, isolation, and maintaining social distance are generally regarded as effective control measures. Existing studies have attempted to incorporate NPIs into infectious disease transmission models to evaluate the effectiveness of different interventions. However, most studies focus on the macro scale, and research on how to effectively implement these measures in micro places is relatively limited. In addition, existing models often rely on historical data for parameter fitting. When a new infectious disease breaks out, the lack of sufficient case data makes it difficult for the model to accurately predict the spread trajectory of the early epidemic. Summary of the invention

[0004] In view of this, the main purpose of the present invention is to provide a method for small-scale group contact risk assessment and infectious disease transmission prediction, which can not only break through the limitations of the traditional age matrix and provide more accurate infectious disease transmission prediction, but also provide accurate situation analysis and intervention measures effect evaluation before or in the early stage of an epidemic. Its flexible and diverse intervention measures simulation function can help managers quickly formulate the best prevention and control strategy, provide more comprehensive and accurate guidance for the prevention and control of epidemics in key places with small spatial scales in closed environments, and protect public health.

[0005] To achieve the above object, the technical solution of the present invention is achieved as follows:

[0006] A method for assessing group contact risk and predicting infectious disease transmission at a small spatial scale, the method comprising:

[0007] Step 1: Using a wearable device to measure the contact behavior data between individuals in a target location within one day; the contact behavior data specifically includes: contact distance, contact start and end timestamps, and contact objects; performing data cleaning on the contact behavior data to obtain pre-processed contact behavior data;

[0008] Step 2: Create an adjacency matrix based on the preprocessed contact behavior data to abstractly represent the contact behavior between each individual; establish an undirected weighted contact network between groups within a small spatial scale;

[0009] Step 3: The ABM model is integrated with the undirected weighted contact network to obtain the ABM model based on the contact network; the risk of each individual is evaluated by the ABM model based on the contact network to obtain the risk assessment result;

[0010] Step 4: By setting the infectious disease transmission parameters, fitting the parameters of the ABM model, and setting one or more initial infected persons, the ABM model based on the contact network simulates the dynamic transmission process of the epidemic, evaluates the risk of infectious disease outbreaks, and obtains the infectious disease outbreak risk assessment results.

[0011] Furthermore, the wearable device includes: a smart bracelet, a smart badge or other sensor devices with near field communication, Bluetooth, RFID or UWB functions.

[0012] Furthermore, the contact distance and the contact start and end timestamps are recorded at a frequency of not less than once per second; when the wearable device is started, the calibration time synchronization mechanism is started to ensure that the timestamps of all wearable devices are recorded in the same time reference system; the contact object is represented by setting a unique identifier ID of both individuals involved in the contact to uniquely distinguish each individual; the unique identifier ID includes: device ID, personal ID or encrypted anonymous identifier.

[0013] By adopting the above technical scheme, the present invention has the following intended effects: the present invention constructs a contact network of small spatial scale by processing the collected contact behavior data. In traditional models, it is usually assumed that the contact between individuals is uniformly mixed, which is difficult to establish in a closed environment. The present invention overcomes this limitation and constructs an undirected weighted contact network based on actual contact behavior, in which each node represents an individual, the edge represents the contact between individuals, and the weight of the edge is determined by the frequency and duration of the contact. By combining the heterogeneity of individual contact behavior and individual-based modeling, i.e., the ABM model, the present invention can simulate the behavior and interaction of individuals in the contact network and predict the spread trajectory of the epidemic between individuals. The ABM model simulates the spread of infectious diseases between individuals by setting individual attributes (such as whether to be vaccinated, whether to be isolated, etc.) and interaction rules. Unlike the traditional model based on the uniform distribution of the entire population, the present invention can reflect the heterogeneity of contact behavior between individuals, thereby improving the prediction accuracy of the dynamics of infectious disease transmission. When a new infectious disease has just broken out, there is often limited case data, and traditional infectious disease models are difficult to provide accurate situation analysis in the early stages. By combining the actual contact network and the ABM model, the present invention can predict the development trend of the epidemic through simulation in the early stage of the epidemic. Through detailed modeling of individual contact behavior and vaccination status, the model can effectively predict the spread speed, infection peak, peak time and duration of the epidemic when case data is limited. Especially in a closed environment, contact between individuals is frequent and the spread of infectious diseases is fast. The contact risk assessment and epidemic spread simulation provided by the present invention can help managers quickly identify high-risk individuals and high-risk activities in the early stages of an outbreak, and take timely and effective intervention measures to prevent the further spread of the epidemic. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0015] Figure 1 A method flow chart of a method for small-scale spatial group contact risk assessment and infectious disease transmission prediction provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0016] The method of the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments of the present invention.

[0017] Example 1: Reference Figure 1 , a method for assessing group contact risk at a small spatial scale and predicting the spread of infectious diseases, the method comprising:

[0018] Step 1: Using a wearable device to measure the contact behavior data between individuals in a target location within one day; the contact behavior data specifically includes: contact distance, contact start and end timestamps, and contact objects; performing data cleaning on the contact behavior data to obtain pre-processed contact behavior data;

[0019] Step 2: Create an adjacency matrix based on the preprocessed contact behavior data to abstractly represent the contact behavior between each individual; establish an undirected weighted contact network between groups within a small spatial scale;

[0020] Step 3: The ABM model is integrated with the undirected weighted contact network to obtain the ABM model based on the contact network; the risk of each individual is evaluated by the ABM model based on the contact network to obtain the risk assessment result;

[0021] Step 4: By setting the infectious disease transmission parameters, fitting the parameters of the ABM model, and setting one or more initial infected persons, the ABM model based on the contact network simulates the dynamic transmission process of the epidemic, evaluates the risk of infectious disease outbreaks, and obtains the infectious disease outbreak risk assessment results.

[0022] Specifically, the specific implementation process of step 2 is to create an adjacency matrix and construct an undirected weighted contact network based on the pre-processed contact behavior data obtained in step 1, providing data support for the subsequent fusion of the ABM model and the prediction of infectious disease transmission. First, it is necessary to define all individuals in the target venue as a node set. Assuming that there are N individuals in the venue who participate in the contact behavior, these individuals can be represented as a node set V = {v1,v2,...,v N Next, initialize an adjacency matrix A of size N×N, the elements of which are ij Used to represent individual v i With v j During initialization, all values ​​of the matrix A are set to 0 by default, indicating that there is no contact between individuals.

[0023] Then, traverse all the pre-processed contact behavior data obtained in step 1, and for each pair of individuals v with contact behavior i and v j , assign values ​​to the matrix according to the contact data. The contact behavior includes information such as contact distance and contact duration. ij The element value can be obtained by a function f(d ij ,t ij ) is calculated based on the contact distance d between individuals. ij and contact time t ij Determine the weighted value. The closer the contact distance and the longer the contact time, the higher the weight value assigned. The following formula can be used to calculate the contact weight:

[0024]

[0025] where d ij For individual v i and v j The contact distance between ij is the contact duration. If the contact distance is very close (e.g., less than 0.1 m) or the contact time is long (e.g., more than 60 minutes), the weight can be further increased to reflect the higher infection risk of these contact behaviors. It should be noted that the adjacency matrix must maintain symmetry because contact behavior is bidirectional. If individual v i and v j If there is contact between them, then A ij =A ji Therefore, when filling the adjacency matrix, it is necessary to ensure that this symmetry condition holds. For long-distance contact (such as more than 10 meters) or short contact time (such as less than 1 second), these abnormal contact behaviors can be excluded and not counted in the adjacency matrix.

[0026] After the adjacency matrix is ​​created, it is further converted into an undirected weighted contact network. Each node in the contact network represents an individual, the edge represents the contact relationship between individuals, and the weight of the edge represents the frequency or intensity of contact between individuals. ij >0, it means that individual v i With v j There is a contact relationship between them. In this case, an undirected edge is added to the contact network. The weight of the edge is A in the adjacency matrix. ij After the contact network is constructed, the key attributes of the network can be calculated to analyze individual contact behavior. The first is the degree of each node, which indicates how many other individuals an individual has contact with, that is, the node v i The degree d(v i ) is the same as v i The number of connected edges. The degree is calculated as:

[0027]

[0028] Next, the weight sum of the node is calculated, which represents the total strength of a certain individual's contact with other individuals. i The weight and formula is:

[0029]

[0030] Among them, w ij is the adjacency matrix A ijThe value of is the contact strength. In addition, the density of the entire contact network can also be calculated, which measures the tightness of the network. The formula for network density is:

[0031]

[0032] Where E represents the number of edges in the network and N represents the total number of nodes.

[0033] During the processing, if the edge weights between some individuals are significantly higher than other edges (such as frequent contact), these edges can be marked so that they can be focused on in subsequent infectious disease transmission simulations. In addition, if some nodes do not have any connecting edges, indicating that these individuals have no contact records, they can be treated as isolated nodes or excluded from subsequent analysis. Finally, the undirected weighted contact network is output in a standard data format. Commonly used formats include adjacency matrix format and edge list format. The adjacency matrix format is saved in a two-dimensional array, which is suitable for subsequent mathematical analysis. The edge list format is in the form of triples (v i ,v j ,w ij ) form, indicating the contact relationship and its weight between individuals. These data will be used for the fusion of ABM models and the prediction analysis of infectious disease spread in the subsequent steps.

[0034] The specific implementation process of step 3 is to integrate the undirected weighted contact network constructed based on step 2 with the ABM model to obtain a dynamic model that can simulate the contact behavior between individuals and the process of infectious disease transmission, and use this model to perform risk assessment and generate contact risk assessment results. To achieve this goal, it is first necessary to construct the subject in the ABM model, that is, the representation of each individual in the model. The ABM model describes the dynamic process of infectious disease transmission by simulating the attributes and behaviors of each individual. Each subject in the model represents an individual in the system, and its attributes include health status, vaccination status, individual susceptibility, etc. There are five types of health status, including susceptible (S), latent (E), obvious infected (I), latent infected (A) and recovered (R). In the initial state, all individuals are susceptible unless a certain proportion of initial infected is set. In this model, the behavioral rules of individuals define how individuals contact other individuals and how their states change. According to the constructed contact network, each individual will interact with its contact object according to the connection relationship in the network. The edge weight in the contact network represents the intensity of contact, which is specifically manifested as the frequency and duration of contact. In each time step, individuals come into contact with other individuals based on the edges in the network. The length of this time step can be set to a day or shorter. Through contact, susceptible individuals may be infected, and the probability of infection is determined by the contact duration and distance. The probability of infection transmission P 感染 It can be calculated by the following formula:

[0035]

[0036] Among them, p trans is the infection rate, t contact is the contact duration, d contact is the contact distance, S is the individual's susceptibility parameter, which comprehensively considers whether the individual has been vaccinated and the protective effect of the vaccine. After the infection, the individual will experience a latent period, during which the individual is not contagious. After the incubation period, the individual may develop into an overtly infected person or a latently infected person. Overtly infected people show symptoms, while latently infected people have no obvious symptoms, but both have the ability to spread. For overtly infected people, they can be isolated according to the model settings. After isolation, the individual will no longer contact other individuals to avoid further transmission. For latently infected people, since they do not show symptoms and are difficult to be discovered, they continue to contact other individuals and become potential high-risk spreaders.

[0037] The role of the contact network is to provide the interaction structure between individuals in the simulation. The edges in the network not only define the contact relationship between individuals, but also define the frequency and intensity of contact through edge weights. According to the structure of the contact network, the model can simulate the interaction between individuals at each time step and track the state changes of each individual. Based on this simulation, the dynamic process of infectious disease transmission can be described in detail. In this process, the model evaluates the contact risk of each individual in the entire transmission process by tracking the centrality of each individual in the contact network. The centrality indicators of individuals include degree centrality, betweenness centrality, and closeness centrality. Degree centrality indicates how many other individuals an individual has contact with, and individuals with high degree centrality often have a higher risk of transmission; betweenness centrality measures the bridge role of individuals in the network. Individuals with high betweenness centrality may become super spreaders and play a key role in the spread of the virus between different groups; closeness centrality indicates the distance between an individual and other individuals. Individuals with high closeness centrality may contact other individuals faster, promoting the rapid spread of the virus.

[0038] Through the dynamic simulation of the model, the contact risk assessment results of each individual can be generated. The contact risk of an individual is determined by his centrality in the contact network and his contact with the infected person. Individuals with high centrality and frequent contact with infected people are marked as high-risk individuals. By analyzing the behavior of these individuals, potential high-risk transmission individuals can be identified in advance. In addition, the model will quantify the contact behavior of all individuals and generate simulation results of the spread of infectious diseases. During the operation of the model, by repeatedly simulating different contact scenarios and transmission scenarios, the contact risk of each individual can be accurately assessed, and the transmission trajectory of infectious diseases in the entire network can be predicted. Through multiple simulations of the model, a list of high-risk individuals can be generated, and different scenarios can be evaluated to provide a basis for formulating intervention measures.

[0039] The specific implementation process of step 4 is to use the ABM model based on the contact network, set the infectious disease transmission parameters, perform model fitting, and simulate the dynamic transmission process of the epidemic by setting the initial infected person, evaluate the risk of infectious disease outbreak, and finally generate the infectious disease outbreak risk assessment results. This process involves the setting of model parameters, the selection of initial conditions, the simulation of the dynamic transmission process, and the evaluation and analysis of the results.

[0040] The first is to set the parameters of infectious disease transmission. In order to accurately simulate the spread of infectious diseases, the key transmission parameters in the model need to be set according to the specific characteristics of infectious diseases. These parameters include the contact transmission rate per unit time, p trans , incubation period, infectious period, differences in infectiousness between overtly infected and latently infected persons, etc. Transmission rate p trans It refers to the probability that a susceptible individual will be infected after an infected person comes into contact with a susceptible individual in a unit of time. This parameter is usually obtained through historical data or epidemiological studies. If there is not enough historical data to support it, simulation can be used to obtain a reasonable parameter range.

[0041] The incubation period and the infectious period are key time parameters in the spread of infectious diseases. The incubation period refers to the period of time after an individual is exposed to the virus and becomes infected but not yet infectious; the infectious period refers to the time during which an infected person can spread the virus during the course of the disease. The infectiousness setting is different for overtly infected and latently infected people. Although latently infected people do not show symptoms, they are still contagious. According to the technical briefing, the infectiousness of latently infected people can be expressed by an infectiousness factor k, which reflects the infectiousness of latently infected people relative to overtly infected people.

[0042] Next is the parameter fitting process of the ABM model. During the fitting process, a time range (such as two weeks) is usually selected, and the model parameters are adjusted based on historical epidemic data so that the model can predict the number of new cases per day within the time range as accurately as possible. The specific fitting method can use the least squares method to calculate the root mean square error (RMSE) between the model prediction value and the actual value. The formula is as follows:

[0043]

[0044] Among them, T i is the number of cases predicted by the model on day i, A i is the actual number of cases, and n is the number of days in the fitting time range. By adjusting the parameters, the RMSE value is gradually reduced, and finally a set of parameters with the smallest error is selected as the best fitting parameters of the model.

[0045] After setting the transmission parameters, you need to set the initial infected person. The initial infected person can be one or more individuals. The number and type of initial infected persons can be flexibly selected according to the size of the venue and the density of the crowd. The selection of the initial infected person will affect the initial spread speed of the epidemic. Therefore, in practical applications, multiple simulations can be performed to select different numbers and types of initial infected persons for scenario analysis. For example, you can set a latent infected person as the initial infected person to simulate the scenario where the latent infected person spreads the virus without knowing it; you can also set multiple explicit infected persons and simulate them at the same time to study the impact of multiple infection sources on the spread speed of the epidemic.

[0046] After setting the initial infected person, the dynamic transmission simulation stage of the epidemic begins. The ABM model advances within a set time step (such as one day). Individuals come into contact with other individuals based on the relationships in the contact network and decide whether to be infected based on the probability of transmission. In each time step, the health status of all individuals will change accordingly: susceptible people may enter the incubation period after contacting infected people, and then become overtly infected or latently infected; after experiencing the infectious period, infected people gradually recover or are isolated and are no longer contagious. As the entire transmission process progresses over time, the model will gradually update the status of each individual, recording the number of new cases and the cumulative number of infections each day.

[0047] During the propagation simulation, the model can output the real-time reproduction number R t , which represents the average number of susceptible people that each infected person can infect in the current time step. The formula for calculating the reproduction number is:

[0048]

[0049] The reproduction number is an important indicator of the epidemic spread, usually R t>1 indicates that the epidemic is spreading, while R t <1 means the epidemic situation has been effectively controlled.

[0050] Through simulation, we can obtain several important epidemic spread indicators, such as infection rate, infection peak, peak time and epidemic duration. The infection rate refers to the proportion of susceptible individuals infected during the entire simulation period; the infection peak refers to the moment when the number of newly infected cases per day reaches the maximum value during the simulation process; the epidemic duration is the length of time from the appearance of the initial infected person to the number of new cases being basically zero.

[0051] Finally, by evaluating the simulation results, the results of the infectious disease outbreak risk assessment can be obtained. The outbreak risk assessment results include the epidemic spread trend, the number of infected people and the changes in the reproduction number under different scenarios. This result helps to analyze and predict the future development of the epidemic, and then provide support for prevention and control decisions. Based on multiple simulations of different initial conditions and parameter combinations, the results can be compared to identify high-risk scenarios for infectious disease outbreaks and potential super spreaders, thereby providing more accurate early warning and decision-making support for responding to epidemic outbreaks.

[0052] Through this detailed implementation of step 4, the model can dynamically simulate the spread of infectious diseases and generate risk assessment results. This process provides a basis for formulating more effective epidemic prevention and control strategies.

[0053] The contact behavior data between individuals are obtained through wearable devices or other data collection methods. The data includes information such as contact distance, contact duration, contact frequency and contact objects between individuals. After the contact data is cleaned and preprocessed, an undirected weighted contact network between individuals in the target place is constructed. At this stage, the edge weights of the contact network can represent the frequency or duration of contact between individuals, and the nodes represent individuals. The construction of the contact network provides a basis for the subsequent contact risk assessment before or in the early stage of the epidemic. The edge weights and node characteristics in the network can reveal the closeness of contact between individuals and which individuals play a key bridge role in the network. After the contact network is constructed, the contact behavior of individuals needs to be evaluated. First, the contact behavior of individuals is quantified by analyzing the contact data. For each individual, the contact risk level can be defined by the contact distance and contact duration. Contact risks are usually divided into three categories: low-risk contact, medium-risk contact and high-risk contact. The specific classification criteria are as follows: Low-risk contact: contact events with a contact distance greater than 2 meters and a contact duration of less than 10 minutes. Medium-risk contact: contact events with a distance between 0.1 meters and 2 meters and a contact duration between 10 minutes and 60 minutes. High-risk contact: contact events with a distance less than 0.1 meters and a contact duration of more than 60 minutes.

[0054] By classifying all contact events of each individual, the risk distribution of the individual's contact behavior can be obtained. At the same time, individuals with high contact frequency and more high-risk contacts will be marked as high-contact risk individuals. In the contact network, in addition to the classification of contact risk, it is also necessary to calculate the importance of each individual in the network. Through social network analysis methods, the centrality of individuals in the contact network can be quantified. Specific indicators include degree centrality, closeness centrality, betweenness centrality and eigenvector centrality. Degree centrality: indicates how many other individuals each individual has contacted with. The higher the degree centrality, the more connections the individual has in the network and has contacted with more individuals. Individuals with high degree centrality may be potential "super spreaders". Closeness centrality: indicates the average contact path length between each individual and other individuals. The higher the closeness centrality, the faster the individual contacts with other individuals. If such individuals are infected, they may quickly spread the virus to more people. Betweenness centrality: indicates the bridge role of individuals in the network. Individuals with high betweenness centrality are usually located between multiple subgroups, connecting different groups of individuals. If such individuals are infected, it may cause the virus to spread between different groups, causing a large-scale spread. Eigenvector centrality: Indicates the strength of the connection between an individual and other important individuals. If an individual has strong connections with multiple important individuals, then the individual has greater influence in the communication network. Through the above centrality analysis, individuals who play a key role in communication in the network can be identified. These individuals may be high-risk spreaders who are the focus of attention in the early stage of the epidemic, and targeted intervention measures should be taken.

[0055] Based on the risk assessment of contact behavior and the analysis of the contact network structure, high-risk individuals and high-risk activities can be identified. High-risk individuals are usually those who have high centrality in the contact network and have close contact with multiple other individuals. These individuals may become potential carriers of virus transmission, especially in the early stage of the epidemic, and these individuals should be closely monitored or isolated in advance. At the same time, high-risk activities can also be identified by analyzing the types of activities in the contact network. For example, certain activities (such as gatherings, long-term indoor interactions) may cause a large number of individuals to come into close contact in a short period of time. Such high-risk activities require early warning and control before or in the early stages of the epidemic. It can be recommended to reduce or cancel high-risk activities, especially in the early stages of the outbreak. In order to assess the potential risk of the epidemic, the transmission process of common infectious diseases in the venue can be simulated based on the constructed contact network and ABM model. By setting one or more initial infected persons (such as latent infected persons or explicit infected persons) and setting relevant parameters (such as infection rate, incubation period, infectious period, etc.) according to the transmission characteristics of the infectious disease, the spread scenario of the infectious disease in the absence of intervention measures can be simulated. The model can simulate the spread trend of the epidemic in the future and predict key indicators such as infection rate, peak infection time, number of infected people and duration of the epidemic. All simulation experiments need to be repeated multiple times, and the median result is taken to ensure the robustness of the simulation results. The simulation results can provide predictions and early warnings for venue managers, so that they can take prevention and control measures in advance, such as strengthening ventilation, reducing the gathering of people, and increasing the frequency of sanitation and disinfection, thereby reducing the risk of an epidemic outbreak.

[0056] By identifying high-risk individuals and high-risk activities, as well as simulating infectious disease transmission scenarios, prevention and control strategies can be formulated in advance. Prevention and control strategies should be based on the following aspects: Management of high-risk individuals: Active monitoring measures can be taken for high-risk individuals in the contact network, especially those with high centrality and high contact frequency, such as regular testing and strengthening personal protection measures. If the initial signs of an epidemic appear, isolation measures can also be taken to prevent them from becoming a source of transmission. Control of high-risk activities: For high-risk activities (such as meetings, group activities, etc.), evaluation and adjustments should be made in advance, and measures such as limiting the number of participants, shortening the activity time, or postponing the activity can be taken. At the same time, ensure that the activity venue is well ventilated and strengthen sanitation and disinfection. Contingency plan formulation: Based on the prediction results of the model, contingency plans can be formulated for different infectious disease scenarios, including possible isolation measures, suspension of classes or work, etc., so as to respond quickly when an epidemic breaks out.

[0057] In applications before or in the early stages of an epidemic, it is first necessary to obtain contact behavior data between individuals through wearable devices. This data includes information such as contact distance, contact duration, and contact objects. Through data cleaning and preprocessing, an undirected weighted contact network in the target location can be constructed. The nodes in the network represent individuals, the edges represent contact behaviors between individuals, and the weight of the edges is the sum of the frequency and duration of contact between individuals. The frequency and duration of contact can be calculated using the formula:

[0058]

[0059] where w ij represents the contact weight between individuals i and j, d ij Indicates the contact distance, represents the duration of each contact, N ij is the number of contacts between individuals i and j.

[0060] After the contact network is constructed, the contact risk can be assessed by contact duration and contact distance. The risk grading of contact behavior can be performed using the following criteria: low-risk contact (contact distance greater than 2 meters, contact duration less than 10 minutes), medium-risk contact (contact distance between 0.1 meters and 2 meters, contact duration between 10 minutes and 60 minutes), and high-risk contact (contact distance less than 0.1 meters, contact duration greater than 60 minutes). For each individual, the risk levels of all contact events are counted to generate their contact risk distribution. Subsequently, the contact network is structurally analyzed and the centrality of each individual is calculated. Degree centrality C d Reflects the number of connections for each individual, the formula is:

[0061]

[0062] Among them A ij represents the elements in the adjacency matrix. When there is contact between individuals i and j, A ij = 1. Closeness centrality C c It is used to measure the distance between an individual and other individuals. The formula is

[0063]

[0064] Where d(v i ,v j ) represents the shortest path distance between individuals i and j. Betweenness centrality C b To measure the role of an individual as a bridge between other individuals, the formula is:

[0065]

[0066] where σ stis the number of all shortest paths from node s to node t, σ st (v) is the number of shortest paths from node s to node t that passes through node v. Through centrality analysis, individuals with high transmission risk in the network can be identified.

[0067] At the same time, the risk of the activity is evaluated. The contact network structure of the activity can be calculated by the network density formula. The network density is the ratio of the actual number of edges to the maximum number of edges that may exist. The formula is:

[0068]

[0069] Where E is the total number of edges in the network and N is the number of nodes in the network. The higher the network density, the more frequent the contacts between individuals in the network, and the higher the risk of infection of the activity.

[0070] In order to further predict the spread of an epidemic, the spread of infectious diseases can be simulated based on the contact network and ABM model. First, set one or more initial infected persons, who can be explicit or latent. By setting parameters such as the transmission rate, incubation period, and infectious period of the infectious disease, the spread of the disease between individuals can be simulated. The probability of infection transmission P 感染 It can be calculated by the following formula:

[0071]

[0072] where p trans is the propagation rate per unit time, t contact is the contact duration, d contact is the contact distance, and S is the individual susceptibility coefficient, taking into account vaccination status and individual immunity differences.

[0073] In order to simulate the dynamic spread of the epidemic, we first need to set the time step (such as one day). In each time step, individuals interact with other individuals based on the edges of the contact network. After contacting an infected person, the susceptible individual S has a certain probability of becoming an infected person. After the incubation period, the infected person may develop into an overtly infected person I or a latently infected person A. Overtly infected people can show symptoms and be isolated, while latently infected people continue to spread the disease without symptoms. The transition of infection status can be described by the following set of equations:

[0074] S→E→I / A→R;

[0075] Among them, S is a susceptible person, E is a latent person, I is an overtly infected person, A is a latently infected person, and R is a recovered person.

[0076] During the simulation, multiple key indicators can be output, such as the number of new cases per day T i , real-time regeneration number Rt Etc. Real-time reproduction number R t The calculation formula is:

[0077]

[0078] When R t >1, indicating that the epidemic is spreading; when R t When <1, it indicates that the epidemic is under control.

[0079] By simulating different scenarios multiple times, different epidemic spread prediction results can be generated, including key indicators such as infection rate, peak infection time, peak number of infected people and epidemic duration. The infection rate refers to the proportion of susceptible individuals infected during the simulation period, and the formula is

[0080]

[0081] The peak infection rate refers to the time point when the number of newly infected people in a single day reaches the highest, and the peak infection rate refers to the maximum number of newly infected people in a single day. The duration of the epidemic is the period from the appearance of the initial infection to the number of new cases approaching zero. Before or in the early stages of an epidemic, through the above analysis and simulation, high-risk individuals and activities can be identified and intervention measures can be formulated in advance.

[0082] Embodiment 2: The wearable device includes: a smart bracelet, a smart badge or other sensor devices with near field communication, Bluetooth, RFID or UWB functions.

[0083] Embodiment 3: The contact distance and the contact start and end timestamps are recorded at a frequency of not less than once per second; when the wearable device is started, the calibration time synchronization mechanism is started to ensure that the timestamps of all wearable devices are recorded in the same time reference system; the contact object is represented by setting a unique identifier ID of both individuals involved in the contact to uniquely distinguish each individual; the unique identifier ID includes: device ID, personal ID or encrypted anonymous identifier.

[0084] It is obvious to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential features of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-restrictive, and the scope of the present invention is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any figure mark in the claims should not be regarded as limiting the claims involved. In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the device claim can also be implemented by one unit or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

Claims

1. A method for assessing group contact risk and predicting infectious disease transmission at a small spatial scale, characterized in that: The method comprises: Step 1: Using a wearable device to measure the contact behavior data between individuals in a target location within one day; the contact behavior data specifically includes: contact distance, contact start and end timestamps, and contact objects; performing data cleaning on the contact behavior data to obtain pre-processed contact behavior data; Step 2: Create an adjacency matrix based on the preprocessed contact behavior data to abstractly represent the contact behavior between each individual; establish an undirected weighted contact network between groups within a small spatial scale; Step 3: The ABM model is integrated with the undirected weighted contact network to obtain the ABM model based on the contact network; the risk of each individual is evaluated by the ABM model based on the contact network to obtain the risk assessment result; Step 4: By setting the infectious disease transmission parameters, fitting the parameters of the ABM model, and setting one or more initial infected persons, the ABM model based on the contact network simulates the dynamic transmission process of the epidemic, evaluates the risk of infectious disease outbreaks, and obtains the infectious disease outbreak risk assessment results.

2. The method for small-scale group contact risk assessment and infectious disease transmission prediction according to claim 1, characterized in that: The wearable device includes: a smart bracelet, a smart badge or other sensor devices with near field communication, Bluetooth, RFID or UWB functions.

3. The method for small-scale group contact risk assessment and infectious disease transmission prediction according to claim 2, characterized in that: The contact distance and contact start and end timestamps are recorded at least once per second. When the wearable device is started, the calibration time synchronization mechanism is started to ensure that the timestamps of all wearable devices are recorded in the same time reference system. The contact object is represented by setting a unique identifier ID of both individuals involved in the contact to uniquely distinguish each individual; The unique identifier ID includes: a device ID, a personal ID, or an encrypted anonymous identifier.

Images