Pre-inquiry method and device based on mapping knowledge domain, medium and equipment

Abstract

The invention provides a pre-inquiry method and device based on a knowledge graph, a medium and equipment. The method comprises the following steps: displaying various diseases which can appear on each part of a body through a front-end page, and taking a disease selected by a user as a first disease; determining a first disease node hit in the knowledge graph; generating at least one round of disease inquiry statement according to an associated node path taking the first disease node as a starting point in the knowledge graph, and providing the disease inquiry statement to a front-end page after each round of disease inquiry statement is generated, so that a user replies to the disease inquiry statement; summarizing comprehensive disease information of the user according to the first disease and reply content of the user for each round of disease inquiry statement; diagnosing the disease of the user according to the comprehensive disease information; and displaying the disease to the user through the front-end page, and providing treatment scheme suggestion information about the disease on the front-end page. The accuracy of disease diagnosis can be improved.

Description

Technical Field

[0001] The present invention relates to the field of current medical consultation technology, and in particular to a knowledge graph-based pre-consultation method and device, medium, and equipment. Background Art

[0002] At present, in the field of pre-consultation, most people rely on large-scale machine learning models for disease diagnosis. However, large-scale machine learning models involve data collection and processing, data labeling to build training sample sets, model training, etc., which are relatively cumbersome and time-consuming, and each step in the process may affect the accuracy of disease diagnosis. Summary of the invention

[0003] In response to at least one of the above technical problems, an embodiment of the present invention provides a pre-diagnosis method, device, medium, and equipment based on a knowledge graph.

[0004] According to the first aspect, the pre-diagnosis method based on the knowledge graph provided by the embodiment of the present invention includes:

[0005] Various diseases that may occur in various parts of the body are displayed on the front-end page, so that the user can select a disease according to his or her own physical condition, and the disease selected by the user is used as the first disease;

[0006] According to the first symptom, determining a first symptom node hit in a pre-constructed knowledge graph;

[0007] At least one round of symptom inquiry statements is generated according to the associated node path starting from the first symptom node in the knowledge graph, and after each round of symptom inquiry statements is generated, the symptom inquiry statements are provided to the front-end page so that the user can reply to the symptom inquiry statements; wherein the knowledge graph is constructed based on multiple diseases, various symptoms of each disease, and symptom inquiry statements designed to distinguish different diseases with the same symptom;

[0008] Summarizing the user's comprehensive symptom information according to the first symptom and the user's reply to each round of symptom inquiry statements;

[0009] Diagnose the user's disease based on the comprehensive disease information;

[0010] The disease is presented to the user via a front-end page, and recommended treatment information for the disease is provided on the front-end page.

[0011] In one embodiment, the knowledge graph includes disease nodes, symptom nodes and symptom query nodes. Each disease node has an association relationship with each symptom node corresponding to various symptoms of the disease corresponding to the disease node, and each symptom node has an association relationship with the symptom query node corresponding to the symptom corresponding to the symptom node.

[0012] In one embodiment, the process of constructing the knowledge graph includes:

[0013] Collecting disease symptom data through the human health clinical knowledge base; wherein the disease symptom data includes various diseases and various symptoms of each disease;

[0014] To distinguish different diseases with the same symptom, a symptom inquiry statement is determined, and the symptom inquiry statement is used as a symptom inquiry statement associated with the symptom; wherein the symptom inquiry statement is an inquiry statement that continues to inquire about the details of the user's symptom in order to distinguish different diseases;

[0015] The code of each disease, each symptom corresponding to the disease, and a symptom inquiry statement associated with each symptom are constructed into a structured information;

[0016] The knowledge graph is constructed based on the multiple structured information obtained by construction; wherein each disease corresponds to a disease node, each symptom corresponds to a symptom node, each symptom inquiry statement corresponds to a symptom inquiry node, each disease node corresponding to a disease is connected to each symptom node corresponding to each symptom corresponding to the disease, and each symptom node corresponding to a symptom is connected to a symptom inquiry node corresponding to a symptom inquiry statement having an associated relationship with the disease.

[0017] In one embodiment, generating at least one round of symptom inquiry statements according to the associated node path starting from the first symptom node in the knowledge graph, and providing the symptom inquiry statements to the front-end page after each round of symptom inquiry statements is generated so that the user can reply to the symptom inquiry statements, includes:

[0018] Determine the reply content of the user to the symptom inquiry statement generated in the i-1th round, and determine the i-th symptom node hit by the reply content in the knowledge graph; wherein i is an integer greater than or equal to 2;

[0019] An i-th disease inquiry statement is determined according to the i-th disease inquiry node associated with the i-th disease node in the knowledge graph, and the i-th disease inquiry statement is provided to a front-end page so that a user can reply to the i-th disease inquiry statement.

[0020] In one embodiment, generating at least one round of symptom inquiry statements according to the associated node path starting from the first symptom node in the knowledge graph, and providing the symptom inquiry statements to the front-end page after each round of symptom inquiry statements is generated so that the user can reply to the symptom inquiry statements, further includes:

[0021] According to each symptom corresponding to each symptom node hit in rounds 1 to i-1, various candidate diseases having each symptom are determined;

[0022] Determine each symptom in each candidate disease, exclude each symptom corresponding to each symptom node that has been hit from each symptom, and obtain candidate symptom;

[0023] determining the number of occurrences of each candidate symptom in each candidate disease;

[0024] Sort each candidate disease in descending order of frequency of occurrence;

[0025] The first N candidate diseases in the sorting results are provided to the front page, so that the user can select them on the front page; wherein N is an integer.

[0026] In one embodiment, determining the reply content of the user to the symptom inquiry statement generated in the i-1th round, and determining the i-th symptom node hit by the reply content in the knowledge graph, includes:

[0027] Determining the user's reply content to the symptom inquiry statement generated in round i-1;

[0028] The i-th disease node hit in the knowledge graph is determined according to the reply content of the user to the disease inquiry statement generated in the i-1 round and the disease selected by the user from the N candidate diseases determined in the i-1 round.

[0029] According to the second aspect, the pre-diagnosis device based on the knowledge graph provided by the embodiment of the present invention includes:

[0030] The first display module is used to display various diseases that may occur in various parts of the body through the front-end page, so that the user can select a disease according to his or her own physical condition, and the disease selected by the user is used as the first disease;

[0031] A first determination module, configured to determine, according to the first symptom, a first symptom node hit in a pre-constructed knowledge graph;

[0032] A multi-round inquiry module, for generating at least one round of symptom inquiry statements according to the associated node path starting from the first symptom node in the knowledge graph, and providing the symptom inquiry statements to the front-end page after each round of symptom inquiry statements are generated, so that the user can reply to the symptom inquiry statements; wherein the knowledge graph is constructed based on multiple diseases, various symptoms of each disease, and symptom inquiry statements designed to distinguish different diseases with the same symptom;

[0033] An information summarizing module, configured to summarize the user's comprehensive symptom information according to the first symptom and the user's reply to each round of symptom inquiry statements;

[0034] A disease diagnosis module, used to diagnose the disease of the user based on the comprehensive symptom information;

[0035] The second display module is used to display the disease to the user through the front-end page and provide treatment plan recommendation information about the disease on the front-end page.

[0036] According to a third aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, enables the computer to execute the method provided in the first aspect.

[0037] According to a fourth aspect, an embodiment of the present invention provides a computing device, including a memory and a processor, wherein the memory stores executable code, and when the processor executes the executable code, the method provided in the first aspect is implemented.

[0038] The pre-diagnosis method, device, medium, and equipment based on the knowledge graph provided by the embodiment of the present invention display various symptoms that can occur in various parts of the body through the front-end page, so that the user selects the symptom according to his own physical condition, and takes the symptom selected by the user as the first symptom, and then determines the first symptom node hit in the knowledge graph, thereby generating at least one round of symptom inquiry statements based on the associated node path starting from the first symptom node, and providing the symptom inquiry statement to the front-end page after each round of symptom inquiry statement is generated, so that the user can reply to the symptom inquiry statement. Through multiple rounds of inquiries, the comprehensive symptom information of the user can be summarized, and then the disease existing in the user can be diagnosed based on the comprehensive symptom information. It can be seen that the embodiment of the present invention does not rely on large models, thereby reducing the dependence on computing resources, effectively reducing the operating cost of the medical system, and avoiding diagnostic bias and uncontrollable risks of large models caused by overfitting of the model, and increasing the knowledge interpretability in the pre-diagnosis process. At the same time, by utilizing the structured characteristics of the knowledge graph, the symptom nodes and symptom inquiry nodes can be quickly located, and the user's answer can be guided through the symptom inquiry statement, thereby significantly improving the efficiency and accuracy of the pre-diagnosis. In addition, through precise question chains and humanized interactive methods, the patient user experience is significantly enhanced, reducing their confusion and waiting time during the pre-consultation process. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 Schematic diagram of the process of a pre-diagnosis method based on a knowledge graph in one embodiment of the present invention;

[0040] Figure 2 This is an overall flow chart of a pre-diagnosis method based on a knowledge graph in one embodiment of the present invention;

[0041] Figure 3 It is a structural block diagram of a pre-diagnosis device based on a knowledge graph in one embodiment of the present invention. DETAILED DESCRIPTION

[0042] In the first aspect, the present invention provides a pre-diagnosis method based on a knowledge graph, see Figure 1 and Figure 2 The method comprises the following steps S110 to S160:

[0043] S110, displaying various diseases that may occur in various parts of the body through the front-end page, so that the user can select a disease according to his or her own physical condition, and the disease selected by the user is used as the first disease;

[0044] That is to say, various symptoms that may occur in various parts of the body are displayed on the front-end page, so that users can choose the symptoms they have according to their physical condition, and the disease selected by the user this time is used as the first disease.

[0045] S120, according to the first symptom, determining a first symptom node hit in a pre-constructed knowledge graph;

[0046] Among them, the knowledge graph may include disease nodes, symptom nodes and symptom inquiry nodes. Each disease node has an association relationship with each symptom node corresponding to various symptoms of the disease corresponding to the disease node, and each symptom node has an association relationship with the symptom inquiry node corresponding to the symptom corresponding to the symptom node.

[0047] In one embodiment, the knowledge graph construction process includes the following steps A11 to A14:

[0048] A11. Collect disease symptom data through the human health clinical knowledge base; wherein the disease symptom data includes various diseases and various symptoms of each disease;

[0049] Among them, the Human Health Clinical Knowledge Base is a professional medical information resource that provides a systematic and authoritative method to collect and analyze disease data. Through the Human Health Clinical Knowledge Base, you can obtain detailed information such as disease classification, symptom description, diagnostic criteria, and treatment plans. Using the Human Health Clinical Knowledge Base, doctors and researchers can quickly retrieve the latest medical research results to provide patients with more accurate diagnosis and treatment recommendations. Therefore, authoritative and comprehensive disease data can be obtained from the Human Health Clinical Knowledge Base.

[0050] A12. To distinguish different diseases with the same symptom, determine a symptom query statement, and use the symptom query statement as a symptom query statement associated with the symptom; wherein the symptom query statement is a query statement that continues to ask the user for symptom details in order to distinguish different diseases;

[0051] It can be seen that according to the manifestations of the same symptom in different diseases, it is necessary to design targeted questions to dig deeper into the details of the symptom in order to diagnose the disease more accurately. The following are some examples:

[0052] (1) Regarding headache symptoms that may occur in different diseases:

[0053] Migraine sufferers: "Are your headaches accompanied by nausea, vomiting, or sensitivity to light?"

[0054] Tension headache sufferers: "Do your headaches feel like your head is being wrapped tightly around your head, and when do they usually occur?"

[0055] (2) For fever symptoms in different diseases:

[0056] Infectious illness: "Is your fever accompanied by chills, sweats, or other signs of infection?"

[0057] Autoimmune disease: "Do you have a persistent fever, joint pain, or other chronic symptoms?"

[0058] (3) For abdominal pain symptoms in different diseases:

[0059] Stomach ulcer: "Does your abdominal pain get better after meals or worse when you're fasting?"

[0060] Cholecystitis: "Is your pain in the upper right abdomen, and is it worse after eating fatty foods?"

[0061] (4) For dyspnea in different diseases:

[0062] Asthma: "Is your breathlessness worse at night or after exercise? Is there wheezing?"

[0063] Heart disease: "Is your breathlessness worse when you lie down and better when you sit up?"

[0064] (5) For rashes in different diseases:

[0065] Hives: "Does your rash appear and disappear quickly, and is it itchy?"

[0066] Lupus: "Does your rash get worse after sun exposure? Do you have a facial butterfly rash?"

[0067] It can be seen that in order to help accurately determine the disease, it is necessary to further inquire about the details of the disease previously selected by the user, so as to design a disease inquiry statement. It can be seen that the disease inquiry statement helps doctors distinguish different diseases by focusing on the specific details of the disease, and provides patients with more accurate diagnosis and treatment suggestions.

[0068] A13. Constructing the code of each disease, each symptom corresponding to the disease, and a symptom inquiry statement associated with each symptom into a structured information;

[0069] Understandably, in medical informatics, organizing diseases, disease symptoms, and disease query statements into structured data is an important task to improve data quality and utilization efficiency. This process involves the use of standardized medical terminology and coding systems, such as ICD-10, the tenth edition of the International Classification of Diseases, and SNOMED CT, the standard for clinical medical terminology, to ensure data accuracy and interoperability.

[0070] For example, for the disease "diabetes", its structured information includes the disease code E11.9, the main symptoms: polyuria, polydipsia, weight loss, and related symptom inquiry statements. The details are as follows:

[0071] Disease name: Diabetes

[0072] Disease code: E11.9

[0073] Symptoms: ["polyuria", "polydipsia", "weight loss"]

[0074] Symptom Questions: [["How many times a day do you urinate?", "At what specific times (e.g., at night or in the morning) have you noticed your urination becoming more frequent recently?", "How much caffeine or other stimulant beverages have you consumed recently?"], ["How much water do you drink a day?", "What medications have you taken recently?", "Is your polydipsia accompanied by other symptoms, such as frequent urination, increased urine volume, or thirst?"], ["Have you lost weight due to changes in your eating or exercise habits?", "Have you recently suffered from stress or anxiety?", "Do you have any recent plans or goals to lose weight?"]]

[0075] Through the above disease query statement, it can be further determined whether the user's disease is diabetes or other diseases that also have the symptoms of "polyuria", "polydipsia" and "weight loss".

[0076] A14. Construct the knowledge graph based on the multiple structured information obtained by construction; wherein each disease corresponds to a disease node, each symptom corresponds to a symptom node, each symptom inquiry statement corresponds to a symptom inquiry node, each disease node corresponding to the disease is connected to each symptom node corresponding to each symptom corresponding to the disease, and each symptom node corresponding to the symptom is connected to the symptom inquiry node corresponding to the symptom inquiry statement having an associated relationship with the disease.

[0077] Specifically, we can first extract disease entities, symptom entities, and symptom inquiry statement entities from the structured information, with each entity corresponding to a node. Then we create associations between the nodes: having symptoms, needing to inquire.

[0078] That is to say, when building a knowledge graph, you first need to define core entities, such as disease entities, symptom entities, and symptom inquiry statement entities. Disease entities usually correspond to specific medical diagnoses, such as "diabetes" or "hypertension", and each disease entity has its own unique attributes, such as etiology, epidemiological characteristics, etc. Symptom entities represent symptoms related to the disease, such as "polyuria" or "polydipsia", which can be direct manifestations of the disease or complications. Symptom inquiry statement entities involve questions that doctors or medical professionals may ask during the diagnosis process, such as "How much caffeine or other stimulating beverages have you consumed recently?" Next, these entities are connected through relationships to form the structure of the knowledge graph. The "has a symptom" relationship connects the disease entity with the symptom entity, indicating that a disease may cause a specific symptom. The "need to inquire" relationship connects the symptom entity with the symptom inquiry statement entity, representing the symptoms that need to be paid attention to and inquired during the diagnosis process. For example, in the knowledge graph, the disease node - diabetes is connected to disease entities such as "polyuria" and "polydipsia". At the same time, these disease entities are connected to specific disease query sentence entities, such as "At what specific time (for example, at night or in the morning) do you find that your recent urination is frequent?"

[0079] S130, generating at least one round of symptom inquiry statements according to the associated node path starting from the first symptom node in the knowledge graph, and providing the symptom inquiry statements to the front-end page after each round of symptom inquiry statements is generated, so that the user can reply to the symptom inquiry statements; wherein the knowledge graph is constructed based on multiple diseases, various symptoms of each disease, and symptom inquiry statements designed to distinguish different diseases with the same symptom;

[0080] In one embodiment, the step of generating at least one round of symptom inquiry statements according to the associated node path starting from the first symptom node in the knowledge graph in S130, and providing the symptom inquiry statements to the front-end page after each round of symptom inquiry statements is generated so that the user can reply to the symptom inquiry statements, may include S131 to S132:

[0081] S131, determining the reply content of the user to the symptom inquiry statement generated in the i-1th round, and determining the i-th symptom node hit by the reply content in the knowledge graph; wherein i is an integer greater than or equal to 2;

[0082] S132. Determine an i-th disease inquiry statement according to the i-th disease inquiry node associated with the i-th disease node in the knowledge graph, and provide the i-th disease inquiry statement to a front-end page so that a user can reply to the i-th disease inquiry statement.

[0083] For example, in the first round, according to the first symptom selected by the user on the front-end page, the first symptom node hit is determined in the knowledge graph, and then according to the first symptom inquiry node associated with the first symptom node, the first symptom inquiry statement corresponding to the first symptom inquiry node is provided to the front-end page, so that the user can reply to the first symptom inquiry statement. In the second round, according to the user's reply content in the first round, the second symptom node hit by the reply content in the knowledge graph can be determined, and then the second symptom inquiry node associated with the second symptom node is determined, and the second symptom inquiry statement corresponding to the second symptom inquiry node is provided to the front-end page, so that the user can reply to the second symptom inquiry statement. By analogy, through multiple rounds of inquiries and user feedback, it is helpful to gradually determine the scope of the disease in the future, and then finally diagnose the disease.

[0084] Furthermore, the step of generating at least one round of symptom inquiry statements according to the associated node path starting from the first symptom node in the knowledge graph in S130, and providing the symptom inquiry statements to the front-end page after each round of symptom inquiry statements is generated so that the user can reply to the symptom inquiry statements, may also include S133 to S137:

[0085] S133, determining various candidate diseases having various symptoms according to various symptoms corresponding to various symptom nodes hit in rounds 1 to i-1;

[0086] S134, determining various symptoms among various candidate diseases, excluding various symptoms corresponding to various symptom nodes that have been hit from the various symptoms, and obtaining candidate symptoms;

[0087] S135, determining the number of occurrences of each candidate symptom in each candidate disease;

[0088] S136, sorting the candidate diseases in descending order of the number of occurrences;

[0089] S137. Provide the first N candidate diseases in the sorting result to the front-end page, so that the user can select them on the front-end page; wherein N is an integer.

[0090] For example, in the first round, various diseases with a first symptom are determined as candidate diseases for this round. Then, various symptoms of each candidate disease are determined, and the first symptom of the various symptoms in each candidate disease is excluded, and the remaining symptoms are candidate symptoms. The number of occurrences of each candidate symptom in each candidate disease is counted, and then the candidate symptoms are sorted in descending order according to the number of occurrences to obtain a sorting result. The top N candidate symptoms in the sorting result are provided to the front-end page, so that the user can further select from these N candidate symptoms, which helps to further narrow the scope of the disease.

[0091] Next, in the second round, various diseases with the first symptom and the second symptom corresponding to the second symptom node are determined as candidate diseases for this round. Then, various symptoms of each candidate disease in this round are determined, and the first symptom and the second symptom in the various symptoms of each candidate disease are excluded, and the remaining symptoms are candidate diseases for this round. The number of occurrences of each candidate symptom in this round in each candidate disease in this round is counted, and then the candidate diseases in this round are sorted in descending order according to the number of occurrences to obtain a sorting result. The top N candidate diseases in the sorting result are provided to the front-end page, so that the user can further select from these N candidate diseases, which helps to further narrow the scope of the disease.

[0092] By analogy, through multiple rounds of recommendation of candidate diseases, users can further select their own diseases, which helps to narrow the scope of disease diagnosis and ultimately diagnose the user's disease.

[0093] Among them, the determination of the reply content of the user to the symptom inquiry statement generated in the i-1th round in S131, and the determination of the i-th symptom node hit by the reply content in the knowledge graph may include A21-A22:

[0094] A21, determining the user's reply content to the symptom inquiry statement generated in round i-1;

[0095] A22. Determine the i-th disease node hit in the knowledge graph based on the user's reply content to the disease inquiry statement generated in the i-1 round and the disease selected by the user from the N candidate diseases determined in the i-1 round.

[0096] It can be seen that when determining the i-th disease node, it is necessary to determine the i-th disease node corresponding to the i-th round based on the user's reply content to the disease inquiry statement generated in the i-1 round and the disease selected by the user from the N candidate diseases determined in the i-1 round.

[0097] S140, summarizing the user's comprehensive symptom information according to the first symptom and the user's reply to each round of symptom inquiry statements;

[0098] S150, diagnosing a disease of the user based on the comprehensive symptom information;

[0099] S160: Display the disease to the user via a front-end page, and provide treatment plan recommendation information about the disease on the front-end page.

[0100] After the initial determination of the first symptom, the user's responses in the above multiple rounds, and even the candidate symptom selected by the user, comprehensive symptom information for the user, i.e., comprehensive symptom information, can be determined. Based on the comprehensive symptom information, one or more diseases of the user can be accurately diagnosed, and the diagnosis results can be displayed on the front-end page, while providing treatment plan recommendations for the user's reference.

[0101] It can be seen that the embodiment of the present invention provides an intelligent pre-consultation solution based on the medical knowledge graph. Since it does not rely on large models, it reduces the dependence on computing resources, effectively reduces the operating cost of the medical system, avoids diagnostic bias caused by model overfitting and the uncontrollable risk of large models, and increases the knowledge interpretability in the pre-consultation process. At the same time, by utilizing the structured characteristics of the knowledge graph, it is possible to quickly locate the symptom nodes and symptom inquiry nodes, and guide users to answer through symptom inquiry statements, thereby significantly improving the efficiency and accuracy of the pre-consultation. In addition, through precise question chains and humanized interactive methods, the user experience of patients is significantly enhanced, reducing their confusion and waiting time during the pre-consultation process. In summary, the embodiment of the present invention not only optimizes the pre-consultation process, but also achieves the multiple goals of resource conservation, efficiency improvement and experience improvement.

[0102] Specifically, the embodiments of the present invention have the following beneficial effects:

[0103] (1) Reduce medical costs: Since the reliance on large models is eliminated, the deployment and maintenance costs of the embodiments of the present invention are relatively low. That is, unlike traditional pre-diagnosis systems, the embodiments of the present invention no longer rely on large machine learning models for symptom collection and diagnosis prediction. This eliminates the need for massive data and complex algorithms, simplifies the system architecture, and reduces operating costs and resource consumption.

[0104] (2) Enhanced diagnostic accuracy: The structured nature of the knowledge graph enables the embodiment of the present invention to quickly locate and expand the associated symptom nodes based on the initial symptom provided by the user, forming a logically coherent inquiry path. In addition, since there is no risk interference from large model illusions and uncertainties, the accuracy of diagnosis can be improved and the misdiagnosis rate can be reduced.

[0105] (3) Improve the efficiency of consultation: Through the logical path in the knowledge graph, redundant questions can be avoided, and the query can be directly jumped to the most relevant disease node, saving time and resources. For common diseases, the embodiment of the present invention can quickly complete the preliminary medical history collection, reduce the waiting time of doctors and patients in the pre-consultation stage, and improve the efficiency of consultation. Moreover, based on the structure of the knowledge graph, the embodiment of the present invention can dynamically adjust the query path according to the user's response, and can also avoid irrelevant or repeated questions, ensuring that the collected information is both comprehensive and refined.

[0106] (4) Support for telemedicine and primary care: The embodiments of the present invention are very suitable for telemedicine environments, and patients can complete the initial medical history collection without face-to-face consultation. This is particularly important for patients living in remote areas or with limited mobility. It also reduces the pressure on hospital outpatient clinics and improves the efficiency of primary care services.

[0107] (5) Promote medical education and research: The knowledge graph itself is a valuable knowledge resource that can be used for medical education and clinical research. It can not only help medical students and primary medical staff become familiar with the symptoms and diagnostic processes of various diseases, but also provide researchers with a structured data set to explore disease patterns and predict trends.

[0108] (6) Adaptability and scalability, easy to update and maintain: The knowledge graph can be continuously expanded and adjusted as medical knowledge develops and updates, ensuring flexibility and adaptability. This means that the method provided by the embodiment of the present invention can continue to evolve to respond to emerging diseases and medical challenges. That is, the structure of the knowledge graph allows the embodiment of the present invention to be updated and maintained more easily. When new medical knowledge emerges, these changes can be reflected by simply modifying the nodes and relationships in the graph without having to retrain complex models.

[0109] (7) Knowledge interpretability: Compared with black-box machine learning models, knowledge graphs have higher transparency and interpretability. Medical personnel can clearly understand the logic and basis behind the inquiry, which helps to improve the credibility and acceptance of the diagnosis.

[0110] (8) Enhanced personalized inquiry: By analyzing the user's specific symptoms and medical history, the system can generate a customized sequence of inquiries to more accurately capture the nuances of an individual's health status and support personalized medical decision-making.

[0111] (9) Improving patient experience: By reducing meaningless waiting time and unnecessary inquiries, the embodiments of the present invention improve the patient's satisfaction and comfort in the pre-consultation stage, making the entire medical process smoother.

[0112] In general, the embodiments of the present invention can reduce the medical costs of pre-consultation, eliminate the drawbacks of large model illusion and uncertainty, improve the quality, efficiency and accessibility of medical services, while reducing the burden on medical workers and providing patients with a more personalized medical experience.

[0113] In a second aspect, an embodiment of the present invention provides a pre-diagnosis device based on a knowledge graph, see Figure 3 , the device 100 comprises:

[0114] The first display module 110 is used to display various diseases that may occur in various parts of the body through the front-end page, so that the user can select a disease according to his or her own physical condition, and the disease selected by the user is used as the first disease;

[0115] A first determination module 120, configured to determine, according to the first symptom, a first symptom node hit in a pre-constructed knowledge graph;

[0116] A multi-round inquiry module 130 is used to generate at least one round of symptom inquiry statements according to the associated node path starting from the first symptom node in the knowledge graph, and provide the symptom inquiry statements to the front-end page after each round of symptom inquiry statements are generated, so that the user can reply to the symptom inquiry statements; wherein the knowledge graph is constructed based on multiple diseases, various symptoms of each disease, and symptom inquiry statements designed to distinguish different diseases with the same symptom;

[0117] An information summarizing module 140 is used to summarize the user's comprehensive symptom information according to the first symptom and the user's reply content to each round of symptom inquiry statements;

[0118] A disease diagnosis module 150, configured to diagnose the user's disease based on the comprehensive symptom information;

[0119] The second display module 160 is used to display the disease to the user through the front-end page and provide treatment plan recommendation information about the disease on the front-end page.

[0120] In one embodiment, the knowledge graph includes disease nodes, symptom nodes and symptom query nodes. Each disease node has an association relationship with each symptom node corresponding to various symptoms of the disease corresponding to the disease node, and each symptom node has an association relationship with the symptom query node corresponding to the symptom corresponding to the symptom node.

[0121] In one embodiment, the apparatus comprises:

[0122] A graph construction module is used to collect disease symptom data through the clinical knowledge base of the People's Health Bureau; wherein the disease symptom data includes various diseases and various symptoms of each disease; in order to distinguish different diseases with the same symptom, a symptom query statement is determined, and the symptom query statement is used as a symptom query statement with an associated relationship with the symptom; wherein the symptom query statement is an inquiry statement that continues to inquire about the details of the user's symptom in order to distinguish different diseases; the code of each disease, the various symptoms corresponding to the disease, and the symptom query statement with an associated relationship with each symptom are constructed into a structured information; according to the multiple structured information obtained by construction, the knowledge graph is constructed; wherein each disease corresponds to a disease node, each symptom corresponds to a symptom node, each symptom query statement corresponds to a symptom query node, each disease node corresponding to the disease is connected to each symptom node corresponding to each symptom corresponding to the disease, and each symptom node corresponding to the symptom is connected to the symptom query node corresponding to the symptom query statement with an associated relationship with the symptom.

[0123] In one embodiment, the multi-round inquiry module includes:

[0124] A first determining unit is used to determine the reply content of the user to the symptom inquiry statement generated in the i-1th round, and determine the i-th symptom node hit by the reply content in the knowledge graph; wherein i is an integer greater than or equal to 2;

[0125] The second determining unit is used to determine the i-th disease inquiry statement according to the i-th disease inquiry node associated with the i-th disease node in the knowledge graph, and provide the i-th disease inquiry statement to the front-end page so that the user can reply to the i-th disease inquiry statement.

[0126] In one embodiment, the multi-round inquiry module further includes:

[0127] A third determination unit is used to determine various candidate diseases having various symptoms according to various symptoms corresponding to various symptom nodes hit in the 1st round to the i-1th round;

[0128] A fourth determination unit is used to determine various symptoms among various candidate diseases, and exclude various symptoms corresponding to various symptom nodes that have been hit from the various symptoms to obtain candidate symptoms;

[0129] A fifth determining unit, configured to determine the number of occurrences of each candidate symptom in each candidate disease;

[0130] The symptom ranking unit is used to sort the candidate symptoms in descending order of the number of occurrences;

[0131] The symptom providing unit is used to provide the first N candidate symptoms in the sorting result to the front page, so that the user can select on the front page; wherein N is an integer.

[0132] In one embodiment, the first determination unit is specifically used to: determine the reply content of the user to the disease inquiry statement generated in the i-1 round; determine the i-th disease node hit in the knowledge graph based on the reply content of the user to the disease inquiry statement generated in the i-1 round and the disease selected by the user from the N candidate diseases determined by the user in the i-1 round.

[0133] It is understandable that the explanation, specific implementation, beneficial effects, examples, etc. of the relevant contents in the device provided in the embodiment of the present invention can be found in the corresponding parts of the method provided in the first aspect, and will not be repeated here.

[0134] In a third aspect, an embodiment of the present invention provides a computer-readable medium, on which computer instructions are stored. When the computer instructions are executed by a processor, the processor executes the method provided in the first aspect.

[0135] Specifically, a system or device equipped with a storage medium can be provided, on which software program code that implements the functions of any of the above-mentioned embodiments is stored, and a computer (or CPU or MPU) of the system or device can read and execute the program code stored in the storage medium.

[0136] In this case, the program code itself read from the storage medium can realize the function of any one of the above-mentioned embodiments, and thus the program code and the storage medium storing the program code constitute a part of the present invention.

[0137] The storage medium embodiments for providing the program code include a floppy disk, a hard disk, a magneto-optical disk, an optical disk (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program code can be downloaded from a server computer by a communication network.

[0138] In addition, it should be clear that the functions of any of the above embodiments can be implemented not only by executing the program code read by the computer, but also by enabling an operating system operating on the computer to complete part or all of the actual operations based on instructions from the program code.

[0139] In addition, it can be understood that the program code read from the storage medium is written to a memory provided in an expansion board inserted into the computer or to a memory provided in an expansion module connected to the computer, and then based on the instructions of the program code, a CPU installed on the expansion board or expansion module is enabled to perform part or all of the actual operations, thereby realizing the functions of any of the above-mentioned embodiments.

[0140] It is understandable that the explanation, specific implementation methods, beneficial effects, examples, etc. of the relevant contents in the computer-readable medium provided in the embodiment of the present invention can be found in the corresponding parts of the method provided in the first aspect, and will not be repeated here.

[0141] In a fourth aspect, an embodiment of the present specification provides a computing device, including a memory and a processor, wherein the memory stores executable code, and when the processor executes the executable code, it implements the method in any embodiment of the specification.

[0142] It is understandable that the explanation, specific implementation, beneficial effects, examples, etc. of the relevant contents in the computing device provided in the embodiment of the present invention can be found in the corresponding parts of the method provided in the first aspect, and will not be repeated here.

[0143] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0144] Those skilled in the art should be aware that in one or more of the above examples, the functions described in the present invention can be implemented by hardware, software, widgets, or any combination thereof. When implemented by software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium.

[0145] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made on the basis of the technical solution of the present invention should be included in the scope of protection of the present invention.

Claims

1. A pre-diagnosis method based on knowledge graph, characterized in that: include: Various diseases that may occur in various parts of the body are displayed on the front-end page, so that the user can select a disease according to his or her own physical condition, and the disease selected by the user is used as the first disease; According to the first symptom, determining a first symptom node hit in a pre-constructed knowledge graph; At least one round of symptom inquiry statements is generated according to the associated node path starting from the first symptom node in the knowledge graph, and after each round of symptom inquiry statements is generated, the symptom inquiry statements are provided to the front-end page so that the user can reply to the symptom inquiry statements; wherein the knowledge graph is constructed based on multiple diseases, various symptoms of each disease, and symptom inquiry statements designed to distinguish different diseases with the same symptom; Summarizing the user's comprehensive symptom information according to the first symptom and the user's reply to each round of symptom inquiry statements; Diagnose the user's disease based on the comprehensive disease information; The disease is presented to the user via a front-end page, and recommended treatment information for the disease is provided on the front-end page.

2. The method according to claim 1, characterized in that The knowledge graph includes disease nodes, symptom nodes and symptom query nodes. Each disease node has an association relationship with each symptom node corresponding to various symptoms of the disease corresponding to the disease node, and each symptom node has an association relationship with the symptom query node corresponding to the symptom corresponding to the symptom node.

3. The method according to claim 2, characterized in that The construction process of the knowledge graph includes: Collecting disease symptom data through the human health clinical knowledge base; wherein the disease symptom data includes various diseases and various symptoms of each disease; To distinguish different diseases with the same symptom, a symptom inquiry statement is determined, and the symptom inquiry statement is used as a symptom inquiry statement associated with the symptom; wherein the symptom inquiry statement is an inquiry statement that continues to inquire about the details of the user's symptom in order to distinguish different diseases; The code of each disease, each symptom corresponding to the disease, and a symptom inquiry statement associated with each symptom are constructed into a structured information; The knowledge graph is constructed based on the multiple structured information obtained by construction; wherein each disease corresponds to a disease node, each symptom corresponds to a symptom node, each symptom inquiry statement corresponds to a symptom inquiry node, each disease node corresponding to a disease is connected to each symptom node corresponding to each symptom corresponding to the disease, and each symptom node corresponding to a symptom is connected to a symptom inquiry node corresponding to a symptom inquiry statement having an associated relationship with the disease.

4. The method according to claim 2, characterized in that: The step of generating at least one round of symptom inquiry statements according to the associated node path starting from the first symptom node in the knowledge graph, and providing the symptom inquiry statements to the front-end page after each round of symptom inquiry statements is generated, so that the user can reply to the symptom inquiry statements, including: Determine the reply content of the user to the symptom inquiry statement generated in the i-1th round, and determine the i-th symptom node hit by the reply content in the knowledge graph; wherein i is an integer greater than or equal to 2; An i-th disease inquiry statement is determined according to the i-th disease inquiry node associated with the i-th disease node in the knowledge graph, and the i-th disease inquiry statement is provided to a front-end page so that a user can reply to the i-th disease inquiry statement.

5. The method according to claim 4, characterized in that The method further comprises generating at least one round of symptom inquiry statements according to the associated node path starting from the first symptom node in the knowledge graph, and providing the symptom inquiry statements to the front-end page after each round of symptom inquiry statements is generated so that the user can reply to the symptom inquiry statements. According to each symptom corresponding to each symptom node hit in rounds 1 to i-1, various candidate diseases having each symptom are determined; Determine each symptom in each candidate disease, exclude each symptom corresponding to each symptom node that has been hit from each symptom, and obtain candidate symptom; determining the number of occurrences of each candidate symptom in each candidate disease; Sort each candidate disease in descending order of frequency of occurrence; The first N candidate diseases in the sorting results are provided to the front page, so that the user can select them on the front page; wherein N is an integer.

6. The method according to claim 5, characterized in that The step of determining the reply content of the user to the symptom inquiry statement generated in the i-1th round, and determining the i-th symptom node hit by the reply content in the knowledge graph, includes: Determining the user's reply content to the symptom inquiry statement generated in round i-1; The i-th disease node hit in the knowledge graph is determined according to the reply content of the user to the disease inquiry statement generated in the i-1 round and the disease selected by the user from the N candidate diseases determined in the i-1 round.

7. A pre-diagnosis device based on knowledge graph, characterized in that: include: The first display module is used to display various diseases that may occur in various parts of the body through the front-end page, so that the user can select a disease according to his or her own physical condition, and the disease selected by the user is used as the first disease; A first determination module, configured to determine, according to the first symptom, a first symptom node hit in a pre-constructed knowledge graph; A multi-round inquiry module, for generating at least one round of symptom inquiry statements according to the associated node path starting from the first symptom node in the knowledge graph, and providing the symptom inquiry statements to the front-end page after each round of symptom inquiry statements are generated, so that the user can reply to the symptom inquiry statements; wherein the knowledge graph is constructed based on multiple diseases, various symptoms of each disease, and symptom inquiry statements designed to distinguish different diseases with the same symptom; An information summarizing module, configured to summarize the user's comprehensive symptom information according to the first symptom and the user's reply to each round of symptom inquiry statements; A disease diagnosis module, used to diagnose the disease of the user based on the comprehensive symptom information; The second display module is used to display the disease to the user through the front-end page and provide treatment plan recommendation information about the disease on the front-end page.

8. The device according to claim 7, characterized in that The knowledge graph includes disease nodes, symptom nodes and symptom query nodes. Each disease node has an association relationship with each symptom node corresponding to various symptoms of the disease corresponding to the disease node, and each symptom node has an association relationship with the symptom query node corresponding to the symptom corresponding to the symptom node.

9. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed in a computer, the computer is caused to execute the method according to any one of claims 1 to 7.

10. A computing device, characterized in that The method comprises a memory and a processor, wherein the memory stores executable codes, and when the processor executes the executable codes, the method according to any one of claims 1 to 7 is implemented.

Images