Infection disease case risk identification method based on large model agent and electronic equipment

Abstract

The application provides a disease case risk identification method based on a large model intelligent agent and an electronic device, and is applied to the field of disease prevention and control. The case risk identification method comprises the following steps: obtaining case information of a plurality of to-be-identified objects; performing keyword matching on the case information and information in a comprehensive feature knowledge base of infectious diseases to obtain primary identification results of the to-be-identified objects; in the case where the primary identification results indicate that a target object in the plurality of to-be-identified objects has a risk of suffering from a disease, obtaining secondary identification results based on the case information of the target object and feature entities of diseases; and in the case where the secondary identification results indicate that the target object has a risk of suffering from a target disease, processing the case information of the target object and the secondary identification results by using an intelligent agent to obtain tertiary identification results.

Description

Technical Field

[0001] This application relates to the field of disease prevention and control, particularly to the field of infectious disease prevention and control technology that integrates artificial intelligence technology, and more specifically, to a method and electronic device for identifying infectious disease case risks based on a large model intelligent agent. Background Technology

[0002] Early and accurate warning of infectious diseases is a core element in blocking the spread of epidemics and reducing public health risks, and is of crucial significance for protecting public health and maintaining social stability. Currently, infectious disease early warning technologies primarily rely on reported infectious disease data after diagnosis (such as confirmed case information) and structured analytical data (such as symptoms, test results, and diagnostic conclusions extracted from electronic medical records) to conduct early warning analysis. By summarizing and analyzing relevant data from existing outbreaks, these technologies provide references for infectious disease prevention and control decisions, playing a fundamental supporting role in routine infectious disease monitoring and control.

[0003] However, existing infectious disease early warning technologies still face numerous unresolved issues in practical applications, failing to meet the demands of early prevention and control of infectious diseases under the new circumstances. Firstly, there is insufficient timeliness in early warning. Existing early warning mechanisms rely primarily on confirmed or diagnosed data "after the outbreak," resulting in a delay in the early stages of patient visits and causing warning signals to be issued later than the initial phase of infection transmission, missing a crucial window for early intervention. Secondly, the accuracy of early warning is low. Current early warning systems rely solely on structured data, often generated from parsing electronic medical records. This parsing process is prone to information loss and semantic bias, leading to poor data quality and consequently affecting the accuracy of early warning decisions. Thirdly, there is excessive computational cost. While using large models to perform full semantic parsing of all patient data can unlock some value from unstructured data, the large model inference process consumes significant computational resources, making it difficult to adapt to the real-time early warning needs of medical institutions at all levels, thus limiting its practicality and scalability. Therefore, existing infectious disease early warning technologies suffer from significant deficiencies in timeliness, accuracy, and computational adaptability, necessitating a method that can overcome these bottlenecks. Summary of the Invention

[0004] In view of this, this application provides a method and electronic device for identifying infectious disease case risks based on a large model intelligent agent.

[0005] One aspect of this application provides a method for identifying infectious disease case risks based on a large model intelligent agent, including: acquiring case information of multiple objects to be identified;

[0006] Keyword matching is performed on the information of each case and the information in the comprehensive knowledge base of infectious disease characteristics to determine whether the case information includes keywords in the comprehensive knowledge base of infectious disease characteristics that indicate the risk of disease, and to obtain the preliminary identification results of each object to be identified.

[0007] When the primary identification result indicates that the target object among multiple objects to be identified has a risk of disease, a secondary identification result is obtained based on the target object's case information and the characteristic entities of each disease; and

[0008] When the secondary identification result indicates that the target object is at risk of having the target disease, the intelligent agent processes the target object's medical information and the secondary identification result to obtain the tertiary identification result indicating whether the target object has the target disease.

[0009] According to embodiments of this application, based on the case information of the target object and the characteristic entities of each disease, a secondary identification result is obtained, including:

[0010] Keyword matching is performed on the case information of the target object and the feature entities of each syndrome disease to determine whether the case information of the target object includes feature entities that indicate each syndrome disease, and the secondary identification result of the syndrome is obtained.

[0011] Based on the target object's case information and the characteristic entities of each infectious disease in the knowledge graph, the probability of the target object contracting each infectious disease is determined, resulting in a secondary identification result for infectious diseases; and

[0012] Based on the results of the secondary identification of syndromes and the secondary identification of infectious diseases, the secondary identification results are obtained.

[0013] According to embodiments of this application, based on the case information of the target object and the feature entities of each infectious disease in the knowledge graph, the probability of the target object suffering from each infectious disease is determined, resulting in a secondary identification result for infectious diseases, including:

[0014] Based on the number of diseases in the infectious disease dataset and the number of diseases involved in each feature entity, the entity weight of each feature entity is determined.

[0015] Based on the feature entities contained in the case information of the target object and the entity weight of each feature entity, determine the object feature weight of the disease features involved in the target object;

[0016] The infectious disease weight of each infectious disease is determined based on the characteristic entities of the disease characteristics involved in each infectious disease and the entity weight of each characteristic entity.

[0017] Based on the feature entities contained in the case information of the target object, the feature entities of the disease characteristics involved in each infectious disease, and the entity weights of each feature entity, the intersection feature weights of the target object for the disease characteristics involved in each infectious disease are determined; and

[0018] Based on the object feature weights, the infectious disease weights of each infectious disease, and the intersection features, the probability of the target object suffering from each infectious disease is determined, and the secondary identification result of infectious diseases is obtained.

[0019] According to embodiments of this application, based on object feature weights, infectious disease weights of each infectious disease, and intersection features, the probability of a target object contracting each infectious disease is determined, resulting in a secondary infectious disease identification result, including:

[0020] If the probability of the target infectious disease is greater than the first probability threshold, the secondary identification result of the infectious disease indicates that the target subject has the target infectious disease;

[0021] If the probability of the target infectious disease is less than the second probability threshold, the secondary infectious disease identification result indicates that the target subject does not have the target infectious disease; and

[0022] If the probability of the target infectious disease is greater than or equal to the second probability threshold and the probability of the target infectious disease is less than or equal to the first probability threshold, the secondary identification result of the infectious disease indicates that the target object is at risk of contracting the target infectious disease.

[0023] According to embodiments of this application, the method further includes:

[0024] Obtain an assessment sample set, which includes multiple assessment samples. Each assessment sample includes sample case information of the assessment subject and sample labels indicating whether the assessment subject has the disease.

[0025] The evaluation results are obtained by evaluating multiple candidate probability thresholds using an evaluation sample set; and

[0026] Based on the evaluation results, a first probability threshold and a second probability threshold are determined from multiple candidate probability thresholds.

[0027] According to embodiments of this application, keyword matching is performed on the case information of the target object and the feature entities of each syndrome disease to determine whether the case information of the target object includes feature entities indicating each syndrome disease, thereby obtaining a secondary syndrome identification result, including:

[0028] Keyword matching was performed between the symptom information in the case information and the symptom feature entities of each syndrome to obtain the secondary identification results of the first syndrome.

[0029] Keyword matching was performed between the physical signs information in the case information and the physical signs and characteristic entities of the syndrome to obtain the secondary identification results of the second syndrome.

[0030] Keyword matching was performed between epidemiological information in case data and epidemiological characteristic entities of the syndrome to obtain the secondary identification results of the third syndrome; and

[0031] Based on the secondary identification results of the first syndrome, the second syndrome, and the third syndrome, the secondary identification results of the syndrome are determined.

[0032] According to embodiments of this application, keyword matching is performed on case information and a comprehensive infectious disease feature knowledge base to determine whether the case information includes keywords in the comprehensive infectious disease feature knowledge base that indicate a risk of disease, thus obtaining preliminary identification results, including:

[0033] The current disease information in the case information is matched with the set of infectious disease symptom features in the comprehensive infectious disease feature knowledge base by keywords to obtain the first identification result;

[0034] The physical examination information in the case information is matched with the set of infectious disease signs and characteristics in the comprehensive knowledge base of infectious diseases by keywords to obtain the second identification result;

[0035] Keyword matching is performed between historical disease information in case data and the set of infectious disease epidemiological features in the comprehensive infectious disease feature knowledge base to obtain a third identification result; and

[0036] Based on the first identification result, the second identification result, and the third identification result, the initial identification result is determined.

[0037] According to embodiments of this application, an intelligent agent processes the case information and secondary identification results of the target object to obtain a tertiary identification result indicating whether the target object suffers from the target disease, including:

[0038] When the secondary identification result indicates that the target object suffers from the target syndrome disease, the syndrome identification agent is used to process the target object's case information and the secondary identification result to obtain the tertiary identification result. The syndrome identification agent is trained using syndrome case samples.

[0039] According to embodiments of this application, an intelligent agent processes the case information and secondary identification results of the target object to obtain a tertiary identification result indicating whether the target object suffers from the target disease, including:

[0040] When the secondary identification result indicates that the target object has the target infectious disease, the infectious disease identification agent is used to process the target object's case information and the secondary identification result to obtain the tertiary identification result. The infectious disease identification agent is trained using infectious disease case samples.

[0041] Another aspect of this application provides an electronic device comprising:

[0042] One or more processors;

[0043] Memory, used to store one or more programs.

[0044] When the one or more programs are executed by the one or more processors, the one or more processors implement the method described above.

[0045] According to the embodiments of this application, the achievable effects include: First, by deeply mining and utilizing case information such as chief complaints and present medical history in the pre-diagnosis stage, the early warning trigger point is moved from "after diagnosis" to the "pre-diagnosis stage" at the initial stage of patient consultation, several hours to several days earlier than existing technologies, thus securing a critical time window for early prevention and control of infectious diseases and significantly improving the timeliness of early warning; Second, by using large-scale model intelligent agents to accurately analyze the deep semantic information of case information, the shortcomings of single-dimensional and incomplete information in structured data are effectively compensated for, avoiding information loss caused by poor data analysis quality, and greatly optimizing the accuracy of early warning; Third, relying on pre-screening modules such as infectious disease case screening and disease risk assessment of characteristic entities, low-risk case data is directly screened out, and large-scale model inference is only initiated for medium- and high-risk data, which not only significantly reduces computing costs and adapts to the computing resources of primary medical institutions, but also simultaneously improves the overall efficiency of risk identification. Attached Figure Description

[0046] The above and other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0047] Figure 1 An exemplary system architecture for identifying infectious disease case risk based on a large model intelligent agent, as described in this application, is shown;

[0048] Figure 2 A flowchart of a method for identifying infectious disease case risk based on a large model intelligent agent according to an embodiment of this application is shown.

[0049] Figure 3 A flowchart illustrating the risk identification process for infectious disease cases with dual identification branches according to an embodiment of this application is shown.

[0050] Figure 4A A flowchart illustrating the infectious disease risk identification process according to an embodiment of this application is shown.

[0051] Figure 4B A flowchart illustrating the risk identification of syndrome diseases according to an embodiment of this application is shown.

[0052] Figure 5 A block diagram of an electronic device suitable for implementing a method for identifying infectious disease cases based on a large model intelligent agent, according to an embodiment of this application, is shown. Detailed Implementation

[0053] The embodiments of this application will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this application. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this application for ease of explanation. However, it will be apparent that one or more embodiments may be implemented without these specific details. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0054] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0055] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0056] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0057] In the embodiments of this application, the collection, updating, analysis, processing, use, transmission, provision, disclosure, and storage of data (e.g., including but not limited to user personal information) comply with relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. In particular, necessary measures have been taken to prevent unauthorized access to user personal information data and to safeguard user personal information security, network security, and national security.

[0058] In the embodiments of this application, the user's authorization or consent was obtained before obtaining or collecting the user's personal information.

[0059] In order to address the core shortcomings of existing infectious disease early warning technologies, such as delayed early warning, high dependence on the quality of structured data leading to low accuracy, and high computational consumption.

[0060] This application provides an embodiment of a method for infectious disease case risk identification based on a large-scale model intelligent agent, comprising: acquiring case information of multiple objects to be identified; performing keyword matching on each case information and information in an infectious disease comprehensive feature knowledge base to determine whether the case information includes keywords in the infectious disease comprehensive feature knowledge base indicating a risk of disease, thereby obtaining a primary identification result for each object to be identified; if the primary identification result indicates that the target object among the multiple objects to be identified has a risk of disease, obtaining a secondary identification result based on the target object's case information and the feature entities of each disease; if the secondary identification result indicates that the target object has a risk of contracting a target disease, processing the target object's case information and the secondary identification result using an intelligent agent to obtain a tertiary identification result indicating whether the target object has contracted the target disease.

[0061] The infectious disease case risk identification method based on large model intelligent agents provided in this application can advance the early warning node to the pre-diagnosis stage, fully integrate case information, and overcome the limitations of traditional technologies that rely solely on post-diagnosis data. At the same time, it can reduce the frequency of large model startup by using technologies such as initial screening of infectious disease case information and secondary recognition of feature entities to screen low-risk data, thereby achieving early identification, accurate assessment and efficient early warning of infectious disease risks. This improves the timeliness and accuracy of early warning, reduces computing power consumption, and provides scientific and practical infectious disease prevention and control technology support for medical institutions and public health departments at all levels, helping to achieve the prevention and control goals of "early detection, early reporting, early isolation and early treatment".

[0062] Compared with traditional infectious disease early warning methods, this application has significant advantages: First, the infectious disease case risk identification method deeply mines and utilizes case information such as the patient's chief complaint and present medical history in the pre-diagnosis stage, shifting the early warning trigger point from "after diagnosis" to the "pre-diagnosis stage" at the initial stage of patient visit. This is several hours to several days earlier than existing technologies, securing a critical time window for early prevention and control of infectious diseases and significantly improving the timeliness of early warning. Second, by using a large-scale model intelligent agent to accurately analyze the deep semantic information of case information, it effectively makes up for the shortcomings of single-dimensional and incomplete information in structured data, avoiding information loss caused by poor data analysis quality and greatly optimizing the accuracy of early warning. Third, relying on pre-screening modules such as infectious disease case screening and disease risk assessment of characteristic entities, low-risk case data is directly screened out, and large-scale model inference is only initiated for medium- and high-risk data. This not only significantly reduces computing costs and adapts to the computing resources of primary healthcare institutions, but also simultaneously improves the overall efficiency of risk identification.

[0063] Figure 1 An exemplary system architecture 100 according to embodiments of this application, which can be applied to a method for identifying infectious disease cases based on a large model intelligent agent, is shown. It should be noted that... Figure 1The examples shown are merely examples of system architectures that can be applied to the embodiments of this application, in order to help those skilled in the art understand the technical content of this application, but do not mean that the embodiments of this application cannot be used in other devices, systems, environments or scenarios.

[0064] like Figure 1 As shown, the system architecture 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 serves as a medium for providing communication links between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired and / or wireless communication links, etc.

[0065] Users can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 via the network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, and / or social media platform software, etc. (for example only).

[0066] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with displays and support web browsing, including but not limited to smartphones, tablets, laptops, and desktop computers.

[0067] In addition, AI glasses 106 can be used to interact with server 105 via network 104 to send messages such as medical records of the object to be identified.

[0068] For example, in a real-world application scenario, the AI ​​glasses 106 can be worn by a doctor who is communicating with a patient seeking medical attention. While communicating, the doctor can collect medical information such as symptoms and signs of the patient through the AI ​​glasses 106 and transmit this information to the server 105 via the network 104.

[0069] Server 105 can be a server that provides various services, such as a backend management server that supports websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103 (this is just an example). The backend management server can analyze and process data such as received user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal devices.

[0070] It should be noted that the infectious disease case risk identification method based on large model intelligent agent provided in this application embodiment can generally be executed by server 105.

[0071] It should be understood that Figure 1 The number of first terminal devices, second terminal devices, third terminal devices, AI glasses, networks, and servers shown in the diagram is merely illustrative. Depending on the implementation requirements, any number of first terminal devices, second terminal devices, third terminal devices, AI glasses, networks, and servers can be included.

[0072] Figure 2 A flowchart of a method for identifying infectious disease case risk based on a large model intelligent agent according to an embodiment of this application is shown.

[0073] like Figure 2 As shown, the method includes operations S201 to S204.

[0074] In operation S201, the case information of each of the multiple objects to be identified is obtained.

[0075] In operation S202, keyword matching is performed on the information of each case and the information in the comprehensive knowledge base of infectious disease characteristics to determine whether the case information includes keywords in the comprehensive knowledge base of infectious disease characteristics that indicate the risk of disease, and to obtain the preliminary identification results of each object to be identified.

[0076] In operation S203, when the primary identification result indicates that the target object among multiple objects to be identified has a risk of disease, a secondary identification result is obtained based on the target object's case information and the characteristic entities of each disease.

[0077] In operation S204, when the secondary identification result indicates that the target object is at risk of having the target disease, the agent processes the target object's medical records and the secondary identification result to obtain a tertiary identification result indicating whether the target object has the target disease.

[0078] Case information can include textual content such as the chief complaint and present medical history before medical treatment. It is used for the self-reported symptom status of the subject to be identified, such as unwell individuals.

[0079] The comprehensive infectious disease characteristic knowledge base can store relevant information about infectious diseases. This knowledge can be obtained by extracting common knowledge such as symptoms, signs, and epidemiological characteristics from specific disease characteristic knowledge bases, stripping away their original disease attribute associations, and then deduplicating, integrating, and standardizing the information. This significantly reduces the amount of knowledge, improves data retrieval efficiency, and can directly serve the convenient application of intelligent agents for the initial screening of infectious disease cases.

[0080] The specialized disease characteristic knowledge base covers the core knowledge such as symptoms, signs, and epidemiological characteristics of various key infectious diseases and syndromes, serving as the foundation of the entire knowledge base system.

[0081] Keywords in case information can be matched with common knowledge such as symptoms, signs, and epidemiological characteristics stored in the comprehensive infectious disease feature knowledge base to obtain preliminary identification results for each object to be identified.

[0082] The initial identification results for each subject can indicate whether there is a risk of disease. If there is a risk, the subject can be designated as the target subject.

[0083] Using this method for initial screening can handle the initial screening of a large number of objects to be identified, improve data processing efficiency, and reduce the consumption of subsequent identification resources.

[0084] Secondary and tertiary identification can be performed when the primary identification results indicate that the target object among multiple objects to be identified is at risk of disease.

[0085] For example, secondary identification includes obtaining secondary identification results based on the case information of the target object and the characteristic entities of each disease.

[0086] It can match the characteristic entities of each disease with case information to accurately determine the probability of a target person having each disease. It can not only identify whether there is a risk of disease, but also accurately locate the type and probability of the disease.

[0087] Secondary identification results can indicate whether the target disease exists and also indicate the confidence level of the identification. If the confidence level is higher than a threshold, the secondary identification result can be used as the target identification result. However, if the secondary identification result indicates that the target object is at risk of having the target disease, an agent processes the target object's medical history and the secondary identification result to obtain a tertiary identification result indicating whether the target object has the target disease.

[0088] Therefore, the large model intelligent agent can be used to identify whether the target object is at risk of contracting an infectious disease. Relying on the powerful semantic understanding capability of the large model, it can deeply analyze the complex diagnostic criteria of the disease and integrate case information such as the chief complaint and present medical history before the diagnosis, and output more accurate and timely judgment results.

[0089] Compared to risk assessment methods that rely solely on large models, the identification method provided in this application can achieve multi-level filtering of infectious disease cases through multi-level collaborative identification, significantly reducing the frequency of large model activation and substantially improving the efficiency and accuracy of case risk identification. Furthermore, it only requires case information of the object to be identified, making it suitable for infectious disease early warning systems and improving warning time. Additionally, the primary identification through keyword matching can handle massive amounts of monitored cases, reducing data processing difficulty while improving identification efficiency.

[0090] The infectious disease case risk identification method based on large model intelligent agents provided in the embodiments of this application can be applied to the identification of various key infectious diseases such as hand-foot-mouth disease, brucellosis, anthrax, and monkeypox, as well as various syndrome diseases such as influenza-like illness, diarrhea syndrome, and fever with thrombocytopenia syndrome.

[0091] According to embodiments of this application, for example, Figure 2 Operation S202, as shown, involves keyword matching between case information and the comprehensive infectious disease feature knowledge base to determine whether the case information includes keywords in the comprehensive infectious disease feature knowledge base indicating a disease risk, thus obtaining a preliminary identification result. This may include: matching the current disease information in the case information with the infectious disease symptom feature set in the comprehensive infectious disease feature knowledge base to obtain a first identification result; matching the physical examination information in the case information with the infectious disease physical sign feature set in the comprehensive infectious disease feature knowledge base to obtain a second identification result; and matching the historical disease information in the case information with the infectious disease epidemiological feature set in the comprehensive infectious disease feature knowledge base to obtain a third identification result. Based on the first, second, and third identification results, the preliminary identification result is determined.

[0092] The comprehensive infectious disease characteristic knowledge base refers to a knowledge base of disease characteristics, which combines the symptom characteristics, physical signs, and epidemiological characteristics of diseases. This comprehensive infectious disease characteristic knowledge base can be used to preliminarily determine whether an object to be identified has the characteristics of an infectious disease.

[0093]

[0094] Where C represents the knowledge set of the comprehensive characteristic knowledge base of infectious diseases, S' represents the characteristic set of infectious disease symptoms, P' represents the characteristic set of infectious disease signs, and E' represents the characteristic set of infectious disease epidemiology.

[0095] Generally, the symptom characteristics of a person to be identified, such as a patient's chief complaint and present illness history, physical signs and characteristics found during physical examination, and epidemiological characteristics of infectious diseases found in past medical history, can all serve as case information for the person to be identified.

[0096] Based on the knowledge set C of the comprehensive infectious disease characteristics knowledge base, collaborative discrimination and hierarchical triage are carried out on three types of structured data (symptoms, signs, and epidemiological history) and four types of unstructured text data (chief complaint, present illness, physical examination, and past medical history) in case information.

[0097] Keyword matching was used to initially identify and triage case characteristics in unstructured text data: for the chief complaint and present medical history text, symptom keywords from the infectious disease symptom feature set S' were matched; for the physical examination text, key physical sign keywords from the infectious disease sign feature set P' were matched; and for the past medical history, epidemiological keywords from the infectious disease epidemiological feature set E' were matched. If any of the above three types of matching detected relevant keywords, the case was determined to be a potential risk case and would directly enter the case risk assessment stage, entering the key infectious disease dual risk assessment agent and the syndrome disease dual risk assessment agent module respectively; if no relevant keywords were detected in any of the three types of matching, it was considered that no infectious disease-related risk was found in the unstructured text, and the case entered the subsequent structured data screening stage.

[0098] The first identification result is obtained by matching the current disease information in the case information with the set of infectious disease symptom features in the comprehensive infectious disease feature knowledge base using keywords. This can be achieved by matching the symptoms S of the object to be identified. p With the set of infectious disease symptom features in the comprehensive knowledge base of infectious disease characteristics Find the intersection of keywords.

[0099] The physical examination information in the case information is matched with the set of infectious disease signs and characteristics in the comprehensive knowledge base of infectious diseases using keywords to obtain a second identification result, which may include the physical signs P of the object to be identified. p With the set of infectious disease signs and characteristics in the comprehensive knowledge base of infectious disease characteristics Find the intersection of keywords.

[0100] The third identification result is obtained by matching keywords between historical disease information in case information and the set of epidemiological features of infectious diseases in the comprehensive knowledge base of infectious disease characteristics. This can include the epidemiological history set E of the object to be identified. p With the set of epidemiological features of infectious diseases in the comprehensive knowledge base of infectious disease characteristics Find the intersection of keywords.

[0101] Based on the first, second, and third identification results, the primary identification result can be determined using the following formula (1).

[0102]

[0103] Formula (1).

[0104] This step calculates the disease risk p(risk of disease). Cases with p(risk of disease) greater than 0 are considered potential risk cases and will proceed directly to secondary identification. Cases with p(risk of disease) = 0 are considered to have no infectious disease association risk, are classified as risk-free cases, and the process ends.

[0105] According to embodiments of this application, primary identification can be performed using feature sets from a comprehensive knowledge base of multiple infectious diseases of different types, thereby improving the identification range and avoiding misidentification.

[0106] According to embodiments of this application, for example, Figure 2 The operation S203 shown, based on the case information of the target object and the feature entities of each disease, obtains a secondary identification result, which may include: performing keyword matching on the case information of the target object and the feature entities of each syndrome disease to determine whether the case information of the target object includes feature entities indicating each syndrome disease, thereby obtaining a secondary identification result for the syndrome; determining the probability of the target object having each infectious disease based on the case information of the target object and the feature entities of each infectious disease in the knowledge graph, thereby obtaining a secondary identification result for the infectious disease; and obtaining a secondary identification result based on the secondary identification result for the syndrome and the secondary identification result for the infectious disease.

[0107] To better link various infectious diseases through symptoms and signs, we can transform the disease knowledge set D based on the disease characteristics knowledge base of a specific infectious disease into an infectious disease knowledge graph.

[0108] Knowledge triples in an infectious disease knowledge graph can include , , etc. Among them, It is an element d of the disease knowledge set D in the knowledge base of specific infectious disease characteristics. i Corresponding infectious diseases, s i p is a symptom element of the symptom feature set S' of infectious disease symptoms. i These are the symptom elements of the symptom set P' of infectious disease symptoms. It is a set of epidemiological characteristics of infectious diseases Epidemiological characteristics.

[0109] By transforming and storing knowledge graphs, rapid dual localization of diseases and various features can be achieved (not only can features be quickly determined based on diseases, but all diseases that match the features can also be quickly located based on the features). This can be applied to the secondary identification stage of key infectious diseases, and disease risk assessment can be achieved through graph reasoning, helping to complete the secondary screening of key infectious diseases with lower computing power.

[0110] Figure 3 A flowchart illustrating the risk identification process for infectious disease cases with dual identification branches according to an embodiment of this application is shown.

[0111] like Figure 3 As shown, keyword matching can be performed on the information of each case and the information in the comprehensive knowledge base of infectious diseases to determine whether the case information includes keywords in the comprehensive knowledge base of infectious diseases that indicate the risk of disease, and to obtain the preliminary identification results of each object to be identified.

[0112] like Figure 3 As shown, when the initial identification results indicate that the target object among multiple objects to be identified has a risk of disease, a multi-branch secondary identification is performed. One branch of secondary identification includes: keyword matching of the target object's case information and the feature entities of each syndrome disease to determine whether the target object's case information includes feature entities indicating each syndrome disease, thus obtaining the syndrome secondary identification result. The other branch of secondary identification includes: based on the target object's case information and the feature entities of each infectious disease in the knowledge graph, determining the probability of the target object having each infectious disease, thus obtaining the infectious disease secondary identification result.

[0113] like Figure 3 As shown, if either the secondary identification result of the syndrome or the secondary identification result of the infectious disease indicates a risk of developing the target disease, then the secondary identification result indicates a risk of developing the target disease. Otherwise, the identification process ends.

[0114] According to the embodiments of this application, by performing recognition in parallel through different branches, the secondary recognition can be highly efficient and have a wide recognition range.

[0115] According to embodiments of this application, for the identification of infectious diseases in secondary identification, based on the case information of the target object and the feature entities of each infectious disease in the knowledge graph, the probability of the target object suffering from each infectious disease is determined to obtain the secondary identification result of infectious diseases. This includes: determining the entity weight of each feature entity based on the number of diseases in the infectious disease dataset and the number of diseases involved in each feature entity; determining the object feature weight of the disease features involved in the target object based on the feature entities contained in the case information of the target object and the entity weight of each feature entity; determining the infectious disease weight of each infectious disease based on the feature entities of the disease features involved in each infectious disease and the entity weight of each feature entity; determining the intersection feature weight of the target object for the disease features involved in each infectious disease based on the feature entities contained in the case information of the target object, the feature entities of the disease features involved in each infectious disease and the entity weight of each feature entity; and determining the probability of the target object suffering from each infectious disease based on the object feature weight, the infectious disease weight of each infectious disease and each intersection feature, to obtain the secondary identification result of infectious diseases.

[0116] Based on the infectious disease knowledge graph, the typicality weights of each symptom characteristic entity, physical sign characteristic entity, and infectious disease epidemiology characteristic entity in the knowledge graph can be calculated to obtain the weights of each entity. For details, please refer to formulas (2), (3), and (4).

[0117] Formula (2).

[0118] Formula (3).

[0119] Formula (4).

[0120] in, , , These are respectively the symptom characteristic entity s, the physical sign characteristic entity p, and the infectious disease epidemiology characteristic entity. Entity weights; B is the set of disease entities in the current knowledge graph; This represents the number of disease entities in the current knowledge graph. , , These represent the symptom feature entity s, the physical sign feature entity p, and the infectious disease epidemiology feature entity in the current knowledge graph. The number of corresponding diseases, It is an infectious disease.

[0121] Based on the feature entities and entity weights of each feature entity contained in the case information of the target object, the object feature weights of the disease features involved in the target object are determined, as shown in formula (5).

[0122] Formula (5).

[0123] Among them, W a It is the sum of entity weights for each characteristic entity involved in the target object's current symptoms, signs, and epidemiological characteristics. It is the sum of the entity weights of the symptom features; It is the sum of the weights of the physical characteristics of the entities; It is the sum of the weights of the entities with epidemic characteristics.

[0124] The infectious disease weight of each infectious disease is determined based on the characteristic entities of the disease characteristics involved in each infectious disease and the entity weight of each characteristic entity. See formula (6).

[0125]

[0126] Formula (6).

[0127] in, yes The sum of the weights of disease-related symptoms, signs, and epidemiological characteristics.

[0128] Based on the feature entities contained in the case information of the target object, the feature entities of the disease characteristics involved in each infectious disease, and the entity weights of each feature entity, the intersection feature weights of the disease characteristics involved in each infectious disease of the target object are determined. This may include calculating the intersection feature weights of the symptoms, signs, and epidemiological characteristics of the target object related to infectious diseases and the symptoms, signs, and epidemiological characteristics of disease b. See formula (7).

[0129]

[0130] Formula (7).

[0131] in, It is the sum of the intersection features of the current infectious disease-related symptoms, signs, and epidemiological characteristics of the target object and the infectious disease-related symptoms, signs, and epidemiological characteristics of disease b.

[0132] Based on object feature weights, infectious disease weights for each infectious disease, and intersection features, the probability of a target object contracting each infectious disease is determined, yielding a secondary infectious disease identification result. This may include calculating the probability that the patient has infectious disease b. For details, please refer to formula (8).

[0133] Formula (8).

[0134] Step 6: Repeat the above formula to calculate the probability of the target object contracting each infectious disease, and obtain the set of the probability of the target object contracting various key infectious diseases.

[0135] According to the embodiments of this application, by performing calculations in the above manner, it is possible to combine big data statistical results, analyze from different dimensions, compare multiple weights, determine the probability of each disease, thereby improving the accuracy of secondary identification.

[0136] According to embodiments of this application, based on object feature weights, infectious disease weights of each infectious disease, and intersection features, the probability of a target object contracting each infectious disease is determined to obtain a secondary infectious disease identification result, including: when the probability of the target infectious disease is greater than a first probability threshold, determining that the secondary infectious disease identification result indicates that the target object has the target infectious disease; when the probability of the target infectious disease is less than a second probability threshold, determining that the secondary infectious disease identification result indicates that the target object does not have the target infectious disease; and when the probability of the target infectious disease is greater than or equal to the second probability threshold and less than or equal to the first probability threshold, determining that the secondary infectious disease identification result indicates that the target object is at risk of contracting the target infectious disease.

[0137] like First probability threshold If the second probability threshold is met, then the target object is considered to possess typical characteristics of infectious disease b. If so, the characteristics of the target object are considered insufficiently typical, constituting an atypical and difficult case of infectious disease b. <First probability threshold If the target is found to be free of infectious disease b, then it is considered that the target is not infected.

[0138] According to the embodiments of this application, the probability is divided into three different thresholds, each corresponding to a different result, which can refine the identification granularity and avoid the problem of misidentification in ambiguous areas.

[0139] According to embodiments of this application, when performing such Figure 2 Prior to operation S203, the method for identifying infectious disease case risk based on a large model agent may further include: obtaining an evaluation sample set, wherein the evaluation sample set includes multiple evaluation samples, and the evaluation samples include sample case information of the evaluation object and sample labels indicating whether the evaluation object is sick; evaluating multiple candidate probability thresholds using the evaluation sample set to obtain evaluation results; and determining a first probability threshold and a second probability threshold from the multiple candidate probability thresholds based on the evaluation results.

[0140] The risk thresholds are respectively the high risk threshold. With low risk threshold High-risk threshold Essentially, it represents the risk value at which the true positive rate is highest, while the low-risk threshold is... It represents the risk of disease at the highest level when a true negative result is achieved.

[0141] The first step is to select 100 clinical cases of each of the n key infectious diseases from real-world case data, resulting in a total of 100n cases.

[0142] The second step is to use, as shown in the example Figure 2 The operations shown are identified one by one.

[0143] The third step is to calculate the high-risk threshold. Let the initial high-risk threshold for disease b be... = 1.0, meaning that infectious disease b is greater than 1.0. If the result is determined to be positive for disease b, calculate the true positive rate for disease b identification under this condition. Then, use a range of 0.1. The values ​​are progressively decreased until the minimum disease risk value for disease b with a true positive rate of 100% is found; this value is the final determined risk value. .

[0144] The fourth step is to calculate the low-risk threshold. Let the initial low-risk threshold for infectious disease b be... = 0.0, meaning that infectious disease b is less than If the result is negative for infectious disease b, calculate the true negative rate for disease b identification under this condition. Then, use a range of 0.1. The values ​​are incremented progressively until the maximum risk value for infectious disease b, where the true negative rate equals 100%, is found. This value is the final determined risk value. value.

[0145] The fifth step is to calculate the low-risk and high-risk thresholds for each disease, following the steps three and four.

[0146] According to embodiments of this application, considering the differences in the number of symptoms and typicality weights for each disease, a threshold should be calculated separately for each disease. This improves the specificity and effectiveness of threshold determination.

[0147] According to embodiments of this application, for example, Figure 2 The operation S204 shown uses an intelligent agent to process the case information and secondary identification results of the target object to obtain a tertiary identification result indicating whether the target object suffers from the target disease. This includes: when the secondary identification result indicates that the target object suffers from the target infectious disease, the infectious disease identification intelligent agent processes the case information and secondary identification results of the target object to obtain the tertiary identification result. The infectious disease identification intelligent agent is trained using infectious disease case samples.

[0148] The infectious disease identification agent is configured with a large infectious disease model. A prompt template can be set for the large infectious disease model to build the infectious disease identification agent.

[0149] For example, "Role: You are a professional physician with in-depth expertise in the diagnosis and treatment of infectious diseases, proficient in various clinical guidelines for infectious diseases, and especially skilled in {{ The diagnosis and differential diagnosis of}} are strictly based on authoritative medical data and do not involve subjective assumptions.

[0150] Task: Based on the following patient medical information, according to {{Y1[ Accurately determine whether it is {{ The function outputs only "Yes" or "No" without any additional explanation.

[0151] Patient medical information: 1. Chief complaint: {{ZS}}; 2. Present illness: {{XBS}}; 3. Physical examination: {{TGJC}}; 4. Past medical history: {{JWS}}.

[0152] in,{{ The variable is contained in the set Y1. Disease name; {{Y1[ The variable is contained in the set Y1. The text is the basis for the diagnosis of the disease.

[0153] Step 2: The diagnostic basis text obtained from the secondary identification results can be filled into the prompt information template to obtain the prompt information. The prompt information and case information are then input into the infectious disease big data model in the infectious disease identification agent to obtain the tertiary identification result. If the tertiary identification result indicates "yes," the patient is considered to have disease b. If the output is "no," the patient is considered to be without disease b. The risk of contracting the disease; if the output is something else, the loop will restart until "yes" or "no" is output.

[0154] The text set used for disease diagnosis is as follows: , , , .

[0155] in, This is a collection of texts used as diagnostic evidence for key infectious diseases. Key infectious diseases The diagnostic basis text; A collection of texts used as diagnostic criteria for syndrome diseases. h is an element of set H i The corresponding syndrome disease name, For syndrome diseases The diagnostic criteria text. X represents a description of the specific diagnostic criteria for each disease.

[0156] by = Take unexplained diarrhea as an example. .

[0157] While current large-scale models possess excellent semantic understanding and logical judgment capabilities, in real-world case identification scenarios for infectious disease prevention and control early warning, their complex model architecture leads to slow inference speeds and excessive computational consumption during concurrent calls, making them unable to support the large-scale model's ability to identify all cases.

[0158] A pre-filtering module was designed to address this pain point. Through initial screening, cases completely lacking any infectious disease-related characteristics are immediately terminated from the subsequent classification process; only cases exhibiting symptoms, signs, or epidemiological characteristics indicating a risk of infectious disease / syndrome are included in the next classification stage. This significantly reduces the number of cases requiring in-depth classification, thereby minimizing the frequency of large model calls and achieving efficient utilization of computing resources.

[0159] Figure 4A A flowchart illustrating the infectious disease risk identification process according to an embodiment of this application is shown.

[0160] like Figure 4A As shown, based on the case information of the target object and the characteristic entities of each infectious disease in the knowledge graph, the probability of the target object suffering from each infectious disease can be determined, and the secondary identification result of infectious diseases can be obtained.

[0161] like Figure 4A As shown, when the secondary identification result indicates that the target object is at risk of contracting the target infectious disease, the target object's case information and the secondary identification result are input into the infectious disease big model in the infectious disease identification agent to obtain the tertiary identification result.

[0162] According to embodiments of this application, for the identification of syndrome diseases in secondary identification, keyword matching is performed on the case information of the target object and the feature entities of each syndrome disease to determine whether the case information of the target object includes feature entities indicating each syndrome disease, thereby obtaining the secondary identification result of the syndrome. This includes: performing keyword matching on the symptom information in the case information with the symptom feature entities of each syndrome disease to obtain a first secondary identification result of the syndrome; performing keyword matching on the physical signs information in the case information with the physical signs feature entities of the syndrome disease to obtain a second secondary identification result of the syndrome; performing keyword matching on the epidemiological information in the case information with the epidemiological feature entities of the syndrome disease to obtain a third secondary identification result of the syndrome; and determining the secondary identification result of the syndrome based on the first secondary identification result of the syndrome, the second secondary identification result of the syndrome, and the third secondary identification result of the syndrome.

[0163] Syndrome risk identification differs from that for key infectious diseases. Syndrome identification can combine logically related conditions such as symptom duration, frequency, and changes in characteristics (e.g., unexplained diarrhea requires "≥3 bowel movements in 24 hours with changes in stool characteristics"). Based on standardized rules in a syndrome topic database, characteristic entities of syndrome symptoms can be extracted. These entities are then used for keyword matching to complete secondary screening, significantly reducing the number of cases required for subsequent deep reasoning and saving computational resources. This process fully adapts to the logical requirements of syndrome determination while achieving efficient utilization of computational resources.

[0164] The following formula (9) can be used for identification.

[0165] Formula (9).

[0166] It represents the set of entities that represent the symptoms of a syndrome. This represents a set of entities that represent the physical characteristics of a syndrome. This represents a set of entities characteristic of the epidemiological history of a syndrome. When... =0, then the patient q is considered to have no risk of developing the disease; when If the value is greater than 0, the patient is considered to have a risk of developing disease q.

[0167] According to embodiments of this application, multiple feature entity sets of different dimensions are used for syndrome identification, thereby improving identification accuracy and effectiveness.

[0168] According to embodiments of this application, for example, Figure 2 The operation S204 shown uses an intelligent agent to process the case information and secondary identification results of the target object to obtain a tertiary identification result indicating whether the target object suffers from the target disease. This includes: when the secondary identification result indicates that the target object suffers from the target syndrome disease, the syndrome identification intelligent agent processes the case information and secondary identification results of the target object to obtain the tertiary identification result. The syndrome identification intelligent agent is trained using syndrome case samples.

[0169] The syndrome recognition agent is configured with a large syndrome model. A prompt template can be set for the large syndrome model to construct the syndrome recognition agent.

[0170] For example, "Role: You are a professional physician with in-depth expertise in the diagnosis and treatment of infectious diseases, proficient in clinical guidelines related to syndromes, and especially skilled in {{ The diagnosis and differential diagnosis of}} are strictly based on authoritative medical data and do not involve subjective assumptions.

[0171] Task: Based on the following patient medical information, according to {{Y2[ Accurately determine whether it is {{ The function outputs only "Yes" or "No" without any additional explanation.

[0172] Patient medical information: 1. Chief complaint: {{ZS}}; 2. Present illness: {{XBS}}; 3. Physical examination: {{TGJC}}; 4. Past medical history: {{JWS}}.

[0173] in,{{ The variable is the name of disease b contained in the set Y2; {{Y2[ The variable is contained in the set Y2. The text is the basis for the diagnosis of the disease.

[0174] Step 2: The diagnostic basis text obtained from the secondary recognition results can be filled into the prompt information template to obtain the prompt information. The prompt information and case information are then input into the syndrome model within the syndrome recognition agent to obtain the tertiary recognition result. If the tertiary recognition result indicates "yes," the patient is considered to have disease b. If the output is "no," the patient is considered to have no disease b. The risk of contracting the disease; if the output is something else, the loop will restart until "yes" or "no" is output.

[0175] Figure 4B A flowchart illustrating the risk identification of syndrome diseases according to an embodiment of this application is shown.

[0176] like Figure 4B As shown, based on the case information of the target object and the symptom feature entities of each syndrome disease in the syndrome topic database, the probability of the target object suffering from each syndrome disease can be determined, and the secondary identification result of the syndrome can be obtained.

[0177] like Figure 4B As shown, when the secondary identification result indicates that the target object has the target syndrome disease, the target object's case information and the secondary identification result are input into the syndrome big model in the syndrome identification agent to obtain the tertiary identification result.

[0178] Figure 5 A block diagram of an electronic device suitable for implementing a method for identifying infectious disease cases based on a large model intelligent agent, according to an embodiment of this application, is shown. Figure 5 The electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0179] like Figure 5 As shown, an electronic device 500 according to an embodiment of this application includes a processor 501, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage portion 508 into a random access memory (RAM) 503. The processor 501 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 501 may also include onboard memory for caching purposes. The processor 501 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of this application.

[0180] RAM 503 stores various programs and data required for the operation of electronic device 500. Processor 501, ROM 502, and RAM 503 are interconnected via bus 504. Processor 501 executes various operations of the method flow according to embodiments of this application by executing programs in ROM 502 and / or RAM 503. It should be noted that the programs may also be stored in one or more memories other than ROM 502 and RAM 503. Processor 501 may also execute various operations of the method flow according to embodiments of this application by executing programs stored in said one or more memories.

[0181] According to embodiments of this application, the electronic device 500 may further include an input / output (I / O) interface 505, which is also connected to a bus 504. The electronic device 500 may also include one or more of the following components connected to the input / output (I / O) interface 505: an input section 506 including a keyboard, mouse, etc.; an output section 507 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 508 including a hard disk, etc.; and a communication section 509 including a network interface card such as a LAN card, modem, etc. The communication section 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the input / output (I / O) interface 505 as needed. A removable medium 511, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 510 as needed so that computer programs read from it can be installed into the storage section 508 as needed.

[0182] According to embodiments of this application, the method flow according to embodiments of this application can be implemented as a computer software program. For example, embodiments of this application include a computer program product comprising a computer program carried on a computer-readable storage medium, the computer program containing program code for performing the methods shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network via communication section 509, and / or installed from removable medium 511. When the computer program is executed by processor 501, it performs the functions defined in the system of embodiments of this application. According to embodiments of this application, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0183] This application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.

[0184] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium. Examples include, but are not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0185] For example, according to embodiments of this application, a computer-readable storage medium may include the ROM 502 and / or RAM 503 described above and / or one or more memories other than ROM 502 and RAM 503.

[0186] Embodiments of this application also include a computer program product comprising a computer program containing program code for performing the methods provided in the embodiments of this application. When the computer program product is run on an electronic device, the program code is used to enable the electronic device to implement the infectious disease case risk identification method based on a large model intelligent agent provided in the embodiments of this application.

[0187] When the computer program is executed by the processor 501, it performs the functions defined in the system / apparatus of this application embodiment. According to the embodiments of this application, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0188] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and may be downloaded and installed via the communication section 509, and / or installed from a removable medium 511. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0189] According to embodiments of this application, program code for executing the computer programs provided in the embodiments of this application can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0190] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions. Those skilled in the art will understand that the features described in the various embodiments of this application can be combined and / or combined in various ways, even if such combinations are not explicitly described in this application. In particular, without departing from the spirit and teachings of this application, the features described in the various embodiments of this application can be combined and / or combined in various ways. All such combinations and / or combinations fall within the scope of this application.

[0191] The embodiments of this application have been described above. However, these embodiments are merely illustrative and not intended to limit the scope of this application. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. Without departing from the scope of this application, those skilled in the art can make various substitutions and modifications, all of which should fall within the scope of this application.

Claims

1. A method for identifying the risk of infectious disease cases based on a large-scale intelligent agent model, characterized in that, The method includes: Obtain the case information of each of the multiple individuals to be identified; Keyword matching is performed on the case information and the information in the comprehensive infectious disease feature knowledge base to determine whether the case information includes keywords in the comprehensive infectious disease feature knowledge base that indicate a risk of disease, and to obtain the preliminary identification results of each of the objects to be identified. When the primary identification result indicates that a target object among the multiple objects to be identified has a risk of disease, a secondary identification result is obtained based on the case information of the target object and the characteristic entities of each disease; and If the secondary identification result indicates that the target object is at risk of having the target disease, the agent processes the target object's medical information and the secondary identification result to obtain a tertiary identification result indicating whether the target object has the target disease.

2. The method according to claim 1, characterized in that, Based on the case information of the target object and the characteristic entities of each disease, a secondary identification result is obtained, including: Keyword matching is performed on the case information of the target object and the feature entities of each syndrome disease to determine whether the case information of the target object includes feature entities indicating each syndrome disease, thereby obtaining the secondary identification result of the syndrome; Based on the case information of the target object and the feature entities of each infectious disease in the knowledge graph, the probability of the target object suffering from each of the infectious diseases is determined, resulting in a secondary infectious disease identification result; and The secondary identification result is obtained based on the secondary identification results of the syndrome and the secondary identification results of the infectious disease.

3. The method according to claim 2, characterized in that, Based on the case information of the target object and the feature entities of each infectious disease in the knowledge graph, the probability of the target object suffering from each of the infectious diseases is determined, and a secondary identification result of the infectious disease is obtained, including: Based on the number of diseases in the infectious disease dataset and the number of diseases involved in the feature entities of each infectious disease, the entity weight of the feature entities of each infectious disease is determined. Based on the feature entities and entity weights of each feature entity contained in the case information of the target object, the object feature weights of the disease features involved in the target object are determined. The infectious disease weight of each infectious disease is determined based on the characteristic entities of the disease characteristics involved in each infectious disease and the entity weight of each characteristic entity. Based on the feature entities contained in the case information of the target object, the feature entities of the disease characteristics involved in each infectious disease, and the entity weights of each feature entity, the intersection feature weights of the target object for each of the disease characteristics involved in the infectious disease are determined; and Based on the object feature weights, the infectious disease weights of each infectious disease, and the intersection features, the probability of the target object contracting each infectious disease is determined, and the secondary identification result of the infectious disease is obtained.

4. The method according to claim 3, characterized in that, Based on the object feature weights, the infectious disease weights of each of the aforementioned infectious diseases, and the intersection features of each of the aforementioned infectious diseases, the probability of the target object contracting each of the aforementioned infectious diseases is determined, resulting in the secondary identification result of the infectious diseases, including: If the probability of the target infectious disease is greater than a first probability threshold, the secondary identification result of the infectious disease indicates that the target object has the target infectious disease; If the probability of the target infectious disease is less than a second probability threshold, the secondary identification result of the infectious disease indicates that the target object does not have the target infectious disease; and If the probability of the target infectious disease is greater than or equal to a second probability threshold and the probability of the target infectious disease is less than or equal to a first probability threshold, the secondary identification result of the infectious disease indicates that the target object is at risk of contracting the target infectious disease.

5. The method according to claim 4, characterized in that, The method further includes: Obtain an assessment sample set, wherein the assessment sample set includes multiple assessment samples, and the assessment samples include sample case information of the assessment object and sample labels indicating whether the assessment object is ill; The evaluation sample set is used to evaluate multiple candidate probability thresholds to obtain evaluation results; and The first probability threshold and the second probability threshold are determined from a plurality of candidate probability thresholds based on the evaluation results.

6. The method according to claim 2, characterized in that, Keyword matching is performed on the case information of the target object and the feature entities of each syndrome disease to determine whether the case information of the target object includes feature entities indicating each syndrome disease, thereby obtaining the secondary syndrome identification result, including: The symptom information in the case information is matched with the disease symptom feature entities of each syndrome by keyword to obtain the secondary identification result of the first syndrome; Keyword matching is performed between the physical signs information in the case information and the physical signs feature entities of the syndrome to obtain the secondary identification result of the second syndrome. Keyword matching was performed between the epidemiological information in the case information and the epidemiological characteristic entities of the syndrome to obtain the secondary identification results of the third syndrome; and Based on the second-level identification results of the first syndrome, the second-level identification results of the second syndrome, and the third-level identification results of the third syndrome, the second-level identification result of the syndrome is determined.

7. The method according to claim 1, characterized in that, Keyword matching is performed between the case information and the comprehensive infectious disease feature knowledge base to determine whether the case information includes keywords in the comprehensive infectious disease feature knowledge base that indicate a risk of disease, thus obtaining preliminary identification results, including: The current disease information in the case information is matched with the set of infectious disease symptom features in the comprehensive infectious disease feature knowledge base by keywords to obtain the first identification result; The physical examination information in the case information is matched with the set of infectious disease signs and characteristics in the comprehensive infectious disease feature knowledge base by keywords to obtain a second identification result; The historical disease information in the case information is matched with the set of infectious disease epidemiological features in the comprehensive infectious disease feature knowledge base using keywords to obtain a third identification result; and Based on the first identification result, the second identification result, and the third identification result, the primary identification result is determined.

8. The method according to claim 1, characterized in that, The intelligent agent processes the case information and the secondary identification results of the target object to obtain a tertiary identification result indicating whether the target object suffers from the target disease, including: When the secondary identification result indicates that the target object suffers from the target syndrome disease, the syndrome identification agent processes the case information of the target object and the secondary identification result to obtain the tertiary identification result. The syndrome identification agent is trained using syndrome case samples.

9. The method according to claim 1, characterized in that, The intelligent agent processes the case information and the secondary identification results of the target object to obtain a tertiary identification result indicating whether the target object suffers from the target disease, including: When the secondary identification result indicates that the target object has a target infectious disease, the infectious disease identification agent processes the case information of the target object and the secondary identification result to obtain the tertiary identification result. The infectious disease identification agent is trained using infectious disease case samples.

10. An electronic device, characterized in that, The electronic device includes: One or more processors; Memory, used to store one or more programs. Wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 9.

Images