Medical information query method based on prompt project and storage medium

Abstract

The invention relates to a medical information query method based on a prompt project and a storage medium, and the method comprises the steps that after a query operation triggered by a target object in a query system is responded, query request information input by the target object is acquired based on the prompt project, and the query request information comprises intention demand information and query category label information; based on the query category label information, at least utilizing a natural language processing (NLP) algorithm to obtain candidate medical information associated with the intention demand information in a target database; and in preset matching rules, determining a target matching rule corresponding to each query operation, screening out target medical information corresponding to the intention demand information from the candidate medical information according to the target matching rule, and taking the target medical information as a query result. Through the method and the device, the problem of low medical information acquisition efficiency and accuracy of a scheme for providing medical information by adopting a large language model in related technologies is solved.

Description

Technical Field

[0001] The present application relates to the cross-technical field of artificial intelligence and medicine, and in particular to a medical information query method and storage medium based on prompt engineering. Background Art

[0002] With the development of medical information technology, the growth rate of medical data far exceeds the learning and absorption capacity of doctors and medical students. Traditional ways of obtaining medical information, including textbooks, medical literature and expert consultation, often have problems such as untimely information updates, long query time, and uneven resource distribution, especially in resource-constrained environments, such as remote areas or emergency situations, where medical information cannot be easily obtained.

[0003] In the related art, although the use of a large language model (LLM) can quickly and accurately process and generate a large amount of medical-related information; however, in the related art, the solution of using a large language model to provide medical information cannot meet the diverse needs of different user groups for obtaining medical information, and the efficiency and accuracy of obtaining medical information are low.

[0004] Regarding the problem of low efficiency and accuracy in obtaining medical information using large language models in related technologies, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present application provide a medical information query method and storage medium based on prompt engineering, as well as an apparatus and an electronic device, so as to at least solve the problem of low efficiency and accuracy in obtaining medical information in the solution of providing medical information using a large language model in the related art.

[0006] In a first aspect, an embodiment of the present application provides a medical information query method based on a prompt project, comprising: after responding to a query operation triggered by a target object in a query system, obtaining query request information input by the target object based on the prompt project, wherein the query request information includes intention demand information and query category label information, and the query category label information is used to characterize one of the following query operations: case screening, medical popularization, information diagnosis, and consultation interaction; based on the query category label information, at least using a natural language processing (NLP) algorithm, obtaining candidate medical information associated with the intention demand information in a target database; in the preset matching rules, determining a target matching rule corresponding to each of the query operations, and according to the target matching rule, screening out the target medical information corresponding to the intention demand information from the candidate medical information, and using the target medical information as the query result.

[0007] In a second aspect, an embodiment of the present application provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the medical information query method based on prompt engineering as described in the first aspect above.

[0008] Compared with the related art, the medical information query method and storage medium based on the prompt engineering provided in the embodiment of the present application adopt, after responding to the query operation triggered by the target object in the query system, to obtain the query request information input by the target object based on the prompt engineering, wherein the query request information includes the intention demand information and the query category label information; based on the query category label information, at least the natural language processing (NLP) algorithm is used to obtain the candidate medical information associated with the intention demand information in the target database; in the preset matching rules, the target matching rule corresponding to each of the query operations is determined, and according to the target matching rule, the target medical information corresponding to the intention demand information is screened out from the candidate medical information, and the target medical information is used as the query result, so as to solve the problem of low efficiency and accuracy of obtaining medical information in the solution of providing medical information using a large language model in the related art, and realize the provision of personalized medical knowledge and suggestions to different user groups, thereby improving the efficiency and accuracy of information acquisition, and being able to conveniently obtain medical information in a resource-constrained environment, thereby improving the accessibility and convenience of medical services, and meeting the diverse needs of users for obtaining medical information.

[0009] Details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more readily apparent. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings: Figure 1 It is a hardware structure block diagram of a terminal of a medical information query method based on prompt engineering according to an embodiment of the present application; Figure 2 is a flow chart of a medical information query method based on prompt engineering according to an embodiment of the present application; Figure 3 It is a structural block diagram of a medical information acquisition device based on prompt engineering according to an embodiment of the present application. DETAILED DESCRIPTION

[0011] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is described and illustrated below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. Based on the embodiments provided in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application. In addition, it can also be understood that although the efforts made in this development process may be complex and lengthy, for ordinary technicians in the field related to the contents disclosed in the present application, some changes such as design, manufacturing or production based on the technical contents disclosed in the present application are only conventional technical means, and should not be understood as insufficient contents disclosed in the present application.

[0012] Reference to "embodiments" in this application means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those of ordinary skill in the art that the embodiments described in this application may be combined with other embodiments without conflict.

[0013] Unless otherwise defined, the technical terms or scientific terms involved in this application should be understood by people with ordinary skills in the technical field to which this application belongs. The words "one", "a", "a", "the" and the like involved in this application do not indicate a quantitative limitation and may represent the singular or plural. The terms "including", "comprising", "having" and any of their variations involved in this application are intended to cover non-exclusive inclusions; for example, a process, method, system, product or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include unlisted steps or units, or may also include other steps or units inherent to these processes, methods, products or devices. The "multiple links" involved in this application refer to links greater than or equal to two. "And / or" describes the association relationship of associated objects, indicating that there may be three relationships, for example, "A and / or B" may represent: A exists alone, A and B exist at the same time, and B exists alone. The terms "first", "second", "third" and the like involved in this application are only used to distinguish similar objects and do not represent a specific ordering of objects.

[0014] Before describing the embodiments of the present application, the related technologies involved in the embodiments of the present application are described as follows: The method embodiment provided in this embodiment can be executed in a terminal, a computer or a similar computing device. Taking running on a terminal as an example, Figure 11 is a hardware structure diagram of a terminal of a medical information query method based on a prompting project in an embodiment of the present application. Figure 1 As shown, the terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data. Optionally, the terminal may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is for illustration only and does not limit the structure of the above terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown.

[0015] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the medical information query method based on prompt engineering in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, to implement the above method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the terminal 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0016] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by the communication provider of the terminal 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0017] This embodiment provides a medical information query method based on prompt engineering running on the above terminal. Figure 2 is a flowchart of a medical information query method based on prompt engineering according to an embodiment of the present application, such as Figure 2 As shown, the process includes the following steps: Step S201, after responding to the query operation triggered by the target object in the query system, obtain the query request information input by the target object based on the prompt project, wherein the query request information includes intention demand information and query category label information, and the query category label information is used to characterize one of the following query operations: case screening, medical popularization, information diagnosis and consultation interaction.

[0018] In this embodiment, the execution subject of the medical information query method of the embodiment of the present application includes but is not limited to the constructed smart medical platform, and integrates a case screening module, a medical popular science module, an information diagnosis module (preliminary diagnosis) and a consultation interaction module; and is used for different target objects respectively; in this embodiment, the target objects include medical practitioners (the smart medical platform provides highly professional and detailed case analysis and diagnosis suggestions, supports multiple rounds of in-depth discussions, and helps doctors make more accurate judgments in complex cases), medical institutions, medical students (the smart medical platform provides learning-oriented case analysis and knowledge point supplementation, combined with the medical popular science module, to help medical students acquire systematic and well-structured knowledge in the learning process), patients (the smart medical platform provides concise and clear health guidance and disease explanations to help patients understand their own health conditions and make preliminary health decisions), and people with health prevention needs; in this embodiment, the operation logic of the smart medical platform starts based on the input of the target object, that is, the target object enters the query system The interface (e.g., web page) of the system (corresponding to the smart medical platform) triggers a query operation (clicking on the integrated modules), and enters query conditions or questions, interacts with the target object based on the prompt project (that is, guides the user to fill in relevant information through a friendly user interface), thereby obtaining the query request information (corresponding to the input data) input by the target object; it can be understood that in the process of the smart medical platform interacting with the target object based on the prompt project, the smart medical platform can also obtain the query category label information selected by the target object (corresponding to determining one of case screening, medical popularization, information diagnosis and consultation interaction) after receiving the corresponding input data, and can also clarify the intention demand information that the target object needs to query and obtain, for example: when the target object is a medical practitioner, when in the process of interacting with the smart medical platform, by entering a symptom description of the target patient, it is expected that the smart medical platform can provide the best treatment plan for the target patient (the intention demand information expected to be obtained by the medical practitioner) Step S202, based on the query category label information, at least using a natural language processing (NLP) algorithm, obtain candidate medical information associated with the intended demand information in the target database.

[0019] In this embodiment, after obtaining the query request information containing the input data input by the target object, the smart medical platform connects with the target database (medical database, medical textbook, medical journal, case database, medical knowledge base, drug knowledge base) through the API interface to obtain relevant case data and medical knowledge, and gather information from medical institutions, medical schools and medical databases, that is, to obtain candidate medical information associated with the intended demand information; in this embodiment, based on the query request information, the NLP technology is used to preliminarily screen out candidate medical information that matches the intended demand information in the target database.

[0020] Step S203, in the preset matching rules, determine the target matching rule corresponding to each query operation, and according to the target matching rule, screen out the target medical information corresponding to the intended demand information from the candidate medical information, and use the target medical information as the query result.

[0021] In this embodiment, after obtaining the candidate medical information, the preset matching rules are used to secondary screen out the target medical information from the candidate medical information. For example, when performing case screening, after obtaining the treatment plans corresponding to the related cases with symptoms similar to those of the patient input by the target object (corresponding to the candidate medical information), the treatment plan with the best effect is selected from the screened treatment plans based on the individual characteristics of the patient. For example, when conducting medical popularization, after the target object inputs a medical keyword, all medical entries containing the medical keyword are preliminarily screened out, and based on the rules for selecting the core data of the drug, irrelevant literature or non-core entries are excluded to obtain the target medical information of medical entries that meet the expected requirements.

[0022] Through the above steps S201 to S203, after responding to the query operation triggered by the target object in the query system, the query request information input by the target object is obtained based on the prompt project, and the query request information includes the intention demand information and the query category label information; based on the query category label information, at least the natural language processing NLP algorithm is used to obtain the candidate medical information associated with the intention demand information in the target database; in the preset matching rules, the target matching rule corresponding to each query operation is determined, and according to the target matching rule, the target medical information corresponding to the intention demand information is screened out from the candidate medical information, and the target medical information is used as the query result, so as to solve the problem of low efficiency and accuracy of obtaining medical information in the solution of providing medical information by using a large language model in the related technology, realize the provision of personalized medical knowledge and suggestions to different user groups, improve the efficiency and accuracy of information acquisition, and can conveniently obtain medical information in a resource-constrained environment, improve the accessibility and convenience of medical services, and meet the diverse needs of users for obtaining medical information.

[0023] It should be noted that the smart medicine platform of the embodiment of the present application provides unified medical information management and services for medical practitioners, medical students and patients by setting up multiple modules including case screening, medical popularization, information diagnosis and consultation interaction; in the present embodiment, through prompt engineering technology, the smart medicine platform can provide personalized medical knowledge and suggestions for different user groups, significantly improving the efficiency and accuracy of information acquisition. In a resource-constrained environment, the target object can obtain medical information anytime and anywhere through the smart medicine platform, which improves the accessibility and convenience of medical services; in the present embodiment, the smart medicine platform of the embodiment of the present application provides a variety of interactive methods for different user groups, and medical practitioners can obtain through the case screening module Detailed case analysis and diagnostic suggestions; medical students can learn basic medical knowledge through the medical popular science module, and have multiple rounds of dialogue with the large language model through the interactive module to consolidate what they have learned; patients can get simple and clear health advice through the preliminary diagnosis module; the smart medicine platform makes the transmission of medical information more in-depth and rich through multi-level interaction and learning experience, meeting the diverse needs of users; in this embodiment, through the prompt engineering technology of the large language model, the smart medicine platform can effectively integrate and utilize medical resources, provide timely and accurate medical advice and diagnosis, and for medical staff and patients with limited resources, the platform can help them make full use of scattered time for learning and health management, thereby improving the efficiency of medical education and medical services.

[0024] In some embodiments, after responding to the query operation triggered by the target object in the query system, the query request information input by the target object is obtained based on the prompting process, which is implemented by the following steps: Step 21, after receiving the query operation triggered by the target object, use the prompt project of the pre-trained large language model to perform multiple rounds of interaction with the corresponding target object, and obtain the target object's response information to the prompt project in each round of interaction.

[0025] Step 22: Determine the intended demand information input by the target object based on the response information.

[0026] Step 23, determining the query category label corresponding to the query operation, and generating query request information according to the response information, intention demand information and query category label corresponding to the target object.

[0027] In this embodiment, the prompt engineering of the large language model is used to generate corresponding prompt sentences, and the corresponding prompt sentences are presented on the interface of the query system. Then, the target object fills in the corresponding response information on the interface according to the prompt sentence. After that, the smart medical platform determines the identity of the target object and the corresponding target information (for example, medical practitioners, medical students) according to the query operation triggered by the target object. In this way, the query category label corresponding to the query operation is determined, and the corresponding intention demand information is determined according to the response information, that is, the medical information that the target object intends to query and obtain is clarified. Finally, the query request information is generated based on the query category label, the intention demand information and the response information; it can be understood that after the target object triggers the corresponding query operation, the query category label will be generated, and the corresponding response information will be classified, and then when the smart medical platform receives the query request information, it can quickly determine the target object currently performing the query request and the intention data that needs to be obtained or queried. Then, the smart medical platform will obtain the relevant medical information from the corresponding target database according to the response information in the query request information.

[0028] After receiving the query operation triggered by the target object in the above steps, the prompt project of the pre-trained large language model is used to perform multiple rounds of interactions with the corresponding target object, and the response information of the target object to the prompt project in each round of interaction is obtained; according to the response information, the intention demand information input by the target object is determined, the query category label corresponding to the query operation is determined, and according to the response information, intention demand information and query category label corresponding to the target object, the query request information is generated to obtain the input data input by the target object, that is, to obtain the query request information.

[0029] In some of the embodiments, based on the query category label information, at least a natural language processing (NLP) algorithm is used to obtain candidate medical information associated with the query requirement information in the target database, which is achieved by the following steps: Step 31, after determining the corresponding query operation according to the query category tag information, obtain corresponding response information from the corresponding query request information.

[0030] In this embodiment, the smart medical platform determines the identity of the target object and the intention demand information corresponding to the response information based on the query category label information representing the corresponding query operation.

[0031] Step 32, using a natural language processing (NLP) algorithm, detect alternative medical feature parameters associated with the first medical feature parameter in the target database, screen the intended medical feature parameters from the alternative feature parameters based on the similarity between the alternative medical feature parameters and the first medical feature parameters, and use the screened intended medical feature parameters as candidate medical information, wherein the first medical feature parameter is generated by preprocessing the response information corresponding to the query operation, and the preprocessing includes at least one of the following: data cleaning, terminology standardization, key feature extraction, normalization, and image enhancement processing.

[0032] In this embodiment, after obtaining the response information input by the target object (as input data), the response information is first preprocessed, and then the corresponding first medical characteristic parameter is generated. For example, the symptom description input by the target object is data cleansed and terminology standardized to generate a symptom code corresponding to the symptom description.

[0033] In this embodiment, after generating the first medical feature data, at least NLP technology is used to detect corresponding alternative medical feature parameters in the target database. For example, the case database is searched for cases similar to the symptom description. For example, the keywords corresponding to the disease code are compared with the symptoms in the case database through NLP technology to find similar cases.

[0034] In this embodiment, after the candidate medical characteristic parameters corresponding to the first medical characteristic parameters are screened out, characteristic analysis and screening are performed according to the corresponding characteristics to obtain the intended medical characteristic parameters.

[0035] In some of the embodiments, detecting candidate medical feature parameters associated with the first medical feature parameter in the target database, and screening intended medical feature parameters from the candidate feature parameters according to the similarity between the candidate medical feature parameters and the first medical feature parameter is achieved by the following steps: Step 41, when it is determined that the query operation is case screening, the BERT natural language processing model is used to perform semantic understanding processing on the first case data corresponding to the response information to obtain the first symptom data.

[0036] Step 42, using the BERT natural language processing model, detects alternative disease data associated with the first disease data in the medical database serving as the target database, and obtains intended disease data from the alternative disease data based on the degree of overlap between the alternative disease data and the first disease data, wherein the degree of overlap is used to characterize the disease similarity of the corresponding disease data.

[0037] Step 43, after determining the second case data associated with the intended symptom data, use a collaborative filtering algorithm based on neighborhood and matrix decomposition to filter out target historical medical information from the historical medical information corresponding to the second case data to obtain candidate medical information, wherein the historical medical information is used to characterize a treatment plan for the symptom in the second case data, and the candidate medical information is used to characterize a preliminary recommended treatment plan for the symptom in the first case data.

[0038] In some embodiments, in the preset matching rules, a target matching rule corresponding to each query operation is determined, and according to the target matching rule, target medical information corresponding to the intended demand information is screened out from the candidate medical information, including the following steps: Step 51, determine the medical user portrait of the first object associated with the first case data, wherein the first object is used to characterize the treatment object corresponding to the first case data, and the medical user portrait includes user characteristics in multiple dimensions, and the user characteristics include one of the following: age, gender, physiological parameters, and medical history parameters.

[0039] Step 52, based on the weighted weights corresponding to all the user features of the first object, a plurality of target historical medical information are screened out from all the target historical medical information corresponding to the candidate medical information.

[0040] Step 53, using the trained treatment rate prediction model, processes the screened multiple target historical medical information, and selects the target historical medical information with the largest predicted treatment rate according to the predicted treatment rate corresponding to each target historical medical information, to obtain the target medical information, wherein the treatment rate prediction model is a prediction model trained based on one of the random forest algorithm and the extreme gradient decision tree.

[0041] In some optional implementations, when the query operation is case screening, the target medical information is obtained through the following steps: Step 1: The target subject inputs the symptom description of the first subject.

[0042] In this embodiment, the symptom description of the first subject includes age, gender, main complaint (such as persistent chest pain, dyspnea), past medical history, and currently taken medications, etc. For example, data describing persistent chest pain of a 30-year-old patient is received.

[0043] Step 2: Preprocess the symptom description.

[0044] In this embodiment, the symptom description is preprocessed, such as data cleaning and terminology standardization; specifically, "received a 30-year-old patient describing persistent chest pain" is truncated to age: extract the patient's age (30 years old); gender: identify according to the gender entered by the patient (male); chief complaint: "persistent chest pain" is standardized into medical terms, for example, "persistent chest pain" may be converted to "persistent chest pain" or "chest pain (persistent)"; redundant information is removed: if there is redundant or irrelevant content in the patient's description (such as repeated descriptions); conversion to standard terminology: use a medical terminology set (such as ICD-10 or SNOMED-CT) to map "chest pain" to a standardized disease code.

[0045] Step 3: Use data visualization and NLP technology to search for cases with similar symptoms to the patient in the case database and return relevant results based on the overlap rate.

[0046] In this embodiment, similar cases are found by identifying the keyword "persistent chest pain" of the symptom and comparing it with the symptoms in the case database through NLP technology. For example, some cases are found in which patients also report "chest pain" or "persistent chest pain", and possible related symptoms, such as whether accompanied by dyspnea, palpitations, sweating, etc.; then, related cases are screened out based on information such as age, gender, duration of symptoms, and the best matching result is returned through overlap (symptom similarity).

[0047] Step 4: Screen out possible relevant treatment options based on the first subject's past medical history and current medication status.

[0048] In this embodiment, based on the medical history and current medication status of the first subject, symptom analysis and screening are performed to determine the most likely cause.

[0049] Step 5: Use the BERT natural language processing model to semantically understand the symptoms and vectorize the input symptoms.

[0050] In this embodiment, One-Hot encoding is used when performing symptom vectorization; in this embodiment, non-numerical data such as disease names, drug names, symptoms, etc. are also encoded. At the same time, each category of disease or drug can be converted into a binary feature through One-Hot encoding (such as "myocardial ischemia" can be [1,0,0], while "diabetes" can be [0,1,0]).

[0051] Step 6: Based on the semantic similarity of the vectorized encoding, symptom matching is performed in the case database to screen out a more accurately matched case set.

[0052] Step 7: Use the treatment history of patients with similar symptoms as a preliminary recommendation, and use a collaborative filtering algorithm (such as a collaborative filtering algorithm based on neighborhood and matrix decomposition) to filter out treatment plans used by patients with similar symptoms.

[0053] In this embodiment, according to the individual characteristics of the first subject, such as age, gender and physical constitution, the weight corresponding to each individual characteristic is adjusted when screening out treatment plans; for example, elderly patients may not be suitable for certain drugs or dosages, so the weight of the age factor is adjusted in the model to make the treatment plan more in line with the needs of a specific population. For example: the interaction risk between the recommended drug and the current drug is detected through a drug knowledge base (such as DrugBank), and potentially conflicting treatment plans are removed; classification models such as random forests or XGBoost are used to predict the success rate of the treatment plan based on patient characteristics and the treatment results of patients with similar symptoms, and the best treatment plan is screened out; and Step 8: Generate personalized treatment recommendations based on the screened treatment options.

[0054] In some of the embodiments, detecting candidate medical feature parameters associated with the first medical feature parameter in the target database, and screening intended medical feature parameters from the candidate feature parameters according to the similarity between the candidate medical feature parameters and the first medical feature parameter, is achieved by the following steps: Step 61, when it is determined that the query operation is medical popular science, search for the target data entry corresponding to the medical keyword corresponding to the response information in the preset vocabulary corresponding to the medical database serving as the target database, wherein the medical database stores a plurality of medical popular science text data, and the medical popular science text data includes at least one medical data entry.

[0055] In this embodiment, after the target object inputs medical keywords as response information, the smart medical platform will determine the target data items associated with the medical keywords based on the input medical keywords and the corresponding vocabulary. For example, when the medical keyword is a drug name, the corresponding associated target data items include: alias, generic name, affiliation, use, source, and manufacturer. In this embodiment, the source of the acquired target medical information is the medical popular science text data stored in the medical database, and each medical popular science text data is composed of a medical data item and a corresponding vocabulary.

[0056] Step 62, using a preset target NLP method, detects target data entries in the medical data entries corresponding to each medical popular science text data, and retrieves medical popular science text data including at least one target data entry as candidate medical popular science text data, and obtains intended medical feature parameters including the candidate medical popular science text data, wherein the target NLP method includes one of the following: forward maximum matching method, reverse maximum matching method, bidirectional scanning method and word-by-word traversal method, and the similarity is determined based on the ratio of the number of target data entries to the number of medical data entries in the medical popular science text data.

[0057] In this embodiment, among all the popular medical science text data stored in the medical database, the target NLP method is used to search, the popular medical science text data containing all the target data items are screened out, and the screened popular medical science text data are used as the intended medical feature parameters.

[0058] In some embodiments, in the preset matching rules, a target matching rule corresponding to each query operation is determined, and according to the target matching rule, target medical information corresponding to the intended demand information is screened out from the candidate medical information, including the following steps: Step 71 , detecting non-core entries among all medical data entries corresponding to the candidate medical popular science text data.

[0059] Step 72, after deleting the candidate medical popular science text data of non-core items detected, the remaining candidate medical popular science text data are used as target medical popular science text data corresponding to the medical keywords, and the target medical popular science text data are used as target medical information corresponding to the intended demand information.

[0060] In some optional implementations, when the query operation is medical science popularization, the target medical information is obtained through the following steps: Step 1: The target subject inputs the keyword (drug name or its alias, generic name, etc.).

[0061] Step 2: After receiving the keyword, the smart medicine platform performs a preliminary search in the medical database using NLP technology based on the keyword to screen out medical popular science text data containing all target data entries of the drug.

[0062] In this embodiment, the target data item includes one of the following: drug ingredients, product instructions, clinical studies, and drug interaction data.

[0063] In this embodiment, the smart medical platform will first extract the drug's trade name, generic name, chemical name and other information from the drug-related target data entries, and confirm whether the drug has a common alias for further matching; in this embodiment, the search includes but is not limited to searching the drug's therapeutic efficacy and mechanism of action, analyzing its effect on specific diseases or symptoms, and sorting out its main indications and secondary indications based on its main components and mechanisms; searching for drug interaction information to find drugs or foods that have a counteracting relationship with it; summarizing the drug's possible synergistic drugs and listing which drugs it can interact with to enhance the efficacy; searching for research literature on the drug to determine its actual clinical application value; combining the drug's mechanism of action, sorting out its unique value in medicine and health management; in this embodiment, the drug's applicable population (such as adults, children, the elderly, pregnant women, etc.) is clarified based on the drug instructions and research data, and the population that is not suitable for use is screened out through side effects and contraindications; extracting detailed information such as recommended dosage, frequency of use, and time of taking medicine from the product instructions or medical advice; comparing whether different methods of taking the drug have a significant impact on the efficacy, and recording adjustment suggestions under special circumstances. Find the known side effects of the drug and summarize the common and rare adverse reactions.

[0064] In some of the optional implementations, the target object inputs a keyword, for example, the target object inputs "aspirin", and the smart medical platform receives the drug name, searches and matches in the medical database, and filters out the entries "Alias: acetylsalicylic acid", "Belongs to: penicillin antibiotics", "Uses: treatment of various infections caused by sensitive bacteria, etc.", "Source: it was originally derived from willow bark, and began to be artificially synthesized in the late 19th century", and "Manufacturer: may include multiple pharmaceutical companies.

[0065] Step 3: From the screened target data items, further select the core information of the drug and exclude irrelevant literature or non-core items.

[0066] In this embodiment, the smart medical platform further screens all relevant target data items that have been screened out, and excludes non-core items, such as excluding irrelevant documents such as "manufacturer" and "shelf life".

[0067] In this embodiment, the core information includes ingredient description and mechanism of action; in this embodiment, in the screened data, various types of information are specifically processed to gradually obtain the required target information. Specifically, the trade name, generic name and chemical name of the drug are extracted from the drug-related entries; confirm whether the drug has common aliases; retrieve the therapeutic efficacy and mechanism of action of the drug, and analyze its effect on specific diseases or symptoms; sort out its main indications and secondary indications based on its main ingredients and mechanisms; retrieve the drug's interaction information to find drugs or foods that have a counteracting relationship with it; summarize the drug's possible synergistic drugs, and list which drugs it can interact with to enhance the efficacy; and search for research literature on the drug. , judge its actual application value in clinical practice; combine the drug's mechanism of action to organize its unique value in medicine and health management; clarify the applicable population of the drug (such as adults, children, the elderly, pregnant women, etc.) according to the drug instructions and research data; screen out people who are not suitable for use through side effects and contraindications; extract detailed information such as recommended dosage, frequency of use, and time of use from product instructions or medical advice; compare whether different methods of taking have a significant effect on the efficacy of the drug, and record adjustment suggestions under special circumstances; find the known side effects of the drug and summarize common and rare adverse reactions; classify and organize side effects according to severity, and attach precautions to avoid adverse reactions.

[0068] Step 4: The smart medical platform organizes and displays this information in the form of text, pictures, videos, etc. for reference by medical personnel or to provide clear medication guidance for patients.

[0069] In this embodiment, when displaying drug information, the smart medicine platform will also intelligently recommend relevant medical articles, research papers or expert lectures based on user portraits to help users gain a deeper understanding of drugs and related knowledge. In this embodiment, the smart medicine platform has an interactive question-and-answer area where users can ask questions about drugs or other medical issues, which will be answered by medical experts or professionally certified volunteers on the smart medicine platform. In this embodiment, the smart medicine platform obtains a large amount of basic medical knowledge from medical textbooks, medical journals, medical databases, etc. The acquired data is structured and NLP technology is used to ensure that relevant information can be accurately matched when users query. Users enter medical knowledge or questions to be queried on the smart medicine platform, and the smart medicine platform will search and match in the medical knowledge base according to the query conditions.

[0070] In some of the embodiments, detecting candidate medical feature parameters associated with the first medical feature parameter in the target database, and screening intended medical feature parameters from the candidate feature parameters according to the similarity between the candidate medical feature parameters and the first medical feature parameter is achieved by the following steps: Step 81, when it is determined that the query operation is information diagnosis and the response information includes medical test data, the pre-processed medical test data is processed using a pre-built convolutional neural network (CNN) to generate a first medical feature parameter, wherein the first medical feature parameter includes diagnostic text data and / or a test image, and the preprocessing includes data cleaning, normalization, key feature extraction, and image enhancement processing.

[0071] Step 82, using a natural language processing (NLP) algorithm based on a knowledge graph, performs a matching analysis on the alternative medical feature parameters corresponding to the historical test data and the first medical feature parameters to determine the similarity between the alternative medical feature parameters and the corresponding first medical feature parameters, wherein the historical test data includes existing medical test data included in a medical database serving as a target database, and the alternative medical feature parameters include first diagnostic text data and / or first test images corresponding to the historical test data.

[0072] Step 83, selecting candidate medical feature parameters whose similarity is greater than a preset threshold, and using the selected candidate medical feature parameters as intended medical feature parameters.

[0073] In some of the embodiments, in the preset matching rules, the target matching rules corresponding to each query operation are determined, and according to the target matching rules, the target medical information corresponding to the intended demand information is screened out from the candidate medical information, including the following steps: obtaining alternative medical feature parameters as intended medical feature parameters, and updating the first medical feature parameters with the corresponding alternative medical feature parameters, and generating target medical information including target medical feature parameters, wherein the target medical feature parameters include target diagnosis text data and / or target detection images.

[0074] In some optional implementations, when the query operation is information diagnosis, the target medical information is obtained through the following steps: Step 1: The target subject inputs a patient's preliminary diagnosis results and related examination data.

[0075] In this embodiment, the preliminary diagnosis result and the related inspection data correspond to the response information.

[0076] Step 2: After receiving the medical test data corresponding to the response information, the smart medical platform verifies and checks to ensure the accuracy and completeness of the medical test data.

[0077] Step 3: The smart medical platform uses the convolutional neural network (CNN) to analyze the preliminary diagnosis results and related examination data, and returns the verification results of the preliminary diagnosis results, as well as further diagnosis suggestions based on the test images (corresponding to the electrocardiogram and blood test results). If the smart medical platform finds that there are errors or omissions in the preliminary diagnosis, it will provide correction suggestions or additional examination suggestions.

[0078] In this embodiment, the smart medicine platform first performs detailed data preprocessing on the data input by the target object, including data cleaning and normalization, and then accurately extracts key features. The extracted features are used to match and analyze case data widely included in the medical database. When the matching similarity is higher than the preset threshold, it indicates that there is a significant similarity between the current case and certain cases in the database. These similar cases can therefore serve as valuable references.

[0079] In this embodiment, the target object can also use the functions provided by the smart medicine platform to scan and detect medical diagnostic images. The smart medicine platform performs image enhancement processing such as edge detection and noise removal on the medical diagnostic images, and compares and analyzes the processed medical diagnostic images with normal images stored in the medical database. Any significantly different areas found in the comparison will be marked in a striking form (such as highlighting, labeling, etc.) and fed back together with the analysis results.

[0080] In some of the embodiments, detecting candidate medical feature parameters associated with the first medical feature parameter in the target database, and screening intended medical feature parameters from the candidate feature parameters according to the similarity between the candidate medical feature parameters and the first medical feature parameter is achieved by the following steps: Step 91, when it is determined that the query operation is a consulting interaction, the BERT natural language processing model is used to perform semantic understanding processing on the preprocessed query text data to obtain semantic features corresponding to the query text, wherein the preprocessing includes data cleaning and the response information includes the query text data.

[0081] Step 92, in a preset medical database, using semantic features and natural language generation (NLG) methods, generate multiple response text data, wherein the alternative medical feature parameters include response text data, and the response text data is used to represent the response to the questions asked by the target object associated with the inquiry text data during the consultation interaction process.

[0082] Step 93, based on the actual demand information determined by parsing the semantic features, calculate the matching degree between each response text data and the actual demand information, and use the matching degree as the corresponding similarity to select at least one response text data as the intended medical feature parameter from multiple response text data.

[0083] In some of the embodiments, in the preset matching rules, the target matching rules corresponding to each query operation are determined, and according to the target matching rules, the target medical information corresponding to the intended demand information is screened out from the candidate medical information, including the following steps: from at least one screened response text data, the response text data with the highest matching degree is selected as the target medical information corresponding to the intended demand information.

[0084] In this embodiment, the smart medicine platform uses natural language processing, prompt engineering technology, large language models, etc. to search and match in the medical knowledge base according to the user's questions and health information, and provides the user with corresponding health advice and preliminary diagnosis; the user enters the consultation question through the chat window, and the smart medicine platform will output the corresponding health advice and preliminary diagnosis results in real time to adjust the user's recent physical condition.

[0085] In some optional implementations, when the query operation is a consultation interaction, the target medical information is obtained through the following steps: Step 1: The target subject inputs a health problem through the smart medicine platform.

[0086] Step 2: After receiving the user's natural language input, the smart medical platform performs semantic understanding and analysis, and uses natural language generation technology and medical databases to generate answers and suggestions for user questions.

[0087] In this embodiment, the smart medicine platform receives text data corresponding to the natural language input by the user, and uses natural language processing (NLP) technology to deeply analyze the user's questions through a large language data model to extract effective information.

[0088] For example, when a user asks "I feel very tired recently, how can I improve my diet and daily routine?", the valid information is accurately identified as "fatigue", "diet" and "daily routine". After receiving this valid information, the platform immediately screens and combines it in the pre-built database to find relevant solutions.

[0089] In this embodiment, the NLP technology processing flow specifically includes: first, carefully screening and preprocessing the natural language text data to remove irrelevant information and retain key content; second, using the NLP model to conduct in-depth analysis of the preprocessed text and convert it into a structured data form. This process can capture the semantic features and emotional tendencies in the text; then, the obtained structured data is parsed, and the user's actual demand information, that is, the intended demand information, is identified and determined through algorithm logic, including but not limited to the user's specific symptoms, areas of concern and desired solutions.

[0090] Step 3: The smart medical platform provides relevant medical knowledge and health tips to help users better understand and manage their health conditions.

[0091] In some of the embodiments, after screening out the target medical information corresponding to the intended demand information, the method further includes: generating target recommendation information based on the target medical information, and displaying the target recommendation information in a preset display format on the interface of the query system, wherein the display format includes one of the following: text display, picture display, video image display.

[0092] This embodiment also provides a medical information acquisition device based on prompt engineering, which is used to implement the above-mentioned embodiments and preferred implementation modes, and will not be repeated hereafter. As used below, the terms "module", "unit", "subunit", etc. can implement a combination of software and / or hardware for a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0093] Figure 3 is a structural block diagram of medical information acquisition based on prompt engineering according to an embodiment of the present application, such as Figure 3 As shown, the device includes an acquisition module 31, a screening module 32 and a processing module 33, wherein: The acquisition module 31 is used to obtain the query request information input by the target object based on the prompt project after responding to the query operation triggered by the target object in the query system, wherein the query request information includes intention demand information and query category label information, and the query category label information is used to characterize one of the following query operations: case screening, medical popularization, information diagnosis and consultation interaction.

[0094] The screening module 32 is coupled to the acquisition module 31 and is used to acquire candidate medical information associated with the intended demand information in the target database based on the query category label information and at least using a natural language processing (NLP) algorithm.

[0095] The processing module 33 is coupled to the screening module 32, and is used to determine the target matching rule corresponding to each query operation in the preset matching rules, and screen out the target medical information corresponding to the intended demand information from the candidate medical information according to the target matching rule, and use the target medical information as the query result.

[0096] In some embodiments, the acquisition module 31 further includes: The interaction unit is used to perform multiple rounds of interactions with the corresponding target object using the prompt project of the pre-trained large language model after receiving the query operation triggered by the target object, and obtain the response information of the target object to the prompt project in each round of interaction.

[0097] The determination unit is coupled to the interaction unit and is used to determine the intended demand information input by the target object based on the response information.

[0098] The generating unit is coupled to the determining unit and is used to determine the query category label corresponding to the query operation, and to generate the query request information according to the response information, the intention demand information and the query category label corresponding to the target object.

[0099] In some embodiments, the screening module 32 further includes: The acquisition unit is used to acquire corresponding response information from the corresponding query request information after determining the corresponding query operation according to the query category label information.

[0100] The selection unit is coupled to the acquisition unit and is used to use a natural language processing (NLP) algorithm to detect alternative medical feature parameters associated with the first medical feature parameter in a target database, screen the intended medical feature parameters from the alternative feature parameters based on the similarity between the alternative medical feature parameters and the first medical feature parameters, and use the screened intended medical feature parameters as candidate medical information, wherein the first medical feature parameter is generated by preprocessing the response information corresponding to the query operation, and the preprocessing includes at least one of the following: data cleaning, terminology standardization, key feature extraction, normalization, and image enhancement processing.

[0101] In some of the embodiments, the selection unit is also used to, when it is determined that the query operation is case screening, use the BERT natural language processing model to perform semantic understanding processing on the first case data corresponding to the response information to obtain the first symptom data; use the BERT natural language processing model to detect alternative symptom data associated with the first symptom data in a medical database serving as a target database, and obtain the intended symptom data from the alternative symptom data based on the overlap between the alternative symptom data and the first symptom data, wherein the overlap is used to characterize the symptom similarity of the corresponding symptom data; after determining the second case data associated with the intended symptom data, use a collaborative filtering algorithm based on neighborhood and matrix decomposition to screen out target historical medical information from the historical medical information corresponding to the second case data to obtain candidate medical information, wherein the historical medical information is used to characterize the treatment plan for the symptom in the second case data, and the candidate medical information is used to characterize the preliminary recommended treatment plan for the symptom in the first case data.

[0102] In some embodiments, the processing module 33 is also used to determine a medical user portrait of a first object associated with the first case data, wherein the first object is used to characterize the treatment object corresponding to the first case data, and the medical user portrait includes user characteristics of multiple dimensions, and the user characteristics include one of the following: age, gender, physiological parameters, and medical history parameters; according to the weighted weights corresponding to all user characteristics of the first object, multiple target historical medical information is screened out from all target historical medical information corresponding to the candidate medical information; using the trained treatment rate prediction model, the screened out multiple target historical medical information is processed, and according to the predicted treatment rate corresponding to each target historical medical information, the target historical medical information with the largest predicted treatment rate is selected to obtain the target medical information, wherein the treatment rate prediction model is a prediction model trained based on one of a random forest algorithm and an extreme gradient decision tree.

[0103] In some of the embodiments, the selection unit is also used to, when it is determined that the query operation is medical popular science, search for a target data entry corresponding to the medical keyword corresponding to the response information in a preset vocabulary corresponding to a medical database serving as a target database, wherein the medical database stores multiple medical popular science text data, and the medical popular science text data includes at least one medical data entry; using a preset target NLP method, detect the target data entry in the medical data entry corresponding to each medical popular science text data, and retrieve the medical popular science text data including at least one target data entry as alternative medical popular science text data, to obtain the intended medical feature parameters including the alternative medical popular science text data, wherein the target NLP method includes one of the following: forward maximum matching method, reverse maximum matching method, bidirectional scanning method, and word-by-word traversal method, and the similarity is determined based on the ratio of the number of target data entries to the number of medical data entries in the medical popular science text data.

[0104] In some of the embodiments, the processing module 33 is also used to detect non-core entries among all medical data entries corresponding to the alternative medical popular science text data; after deleting the alternative medical popular science text data with detected non-core entries, the remaining alternative medical popular science text data is used as the target medical popular science text data corresponding to the medical keywords, and the target medical popular science text data is used as the target medical information corresponding to the intended demand information.

[0105] In some of the embodiments, the selection unit is also used to, when it is determined that the query operation is information diagnosis and the response information includes medical test data, use a pre-built convolutional neural network (CNN) to process the pre-processed medical test data to generate a first medical feature parameter, wherein the first medical feature parameter includes diagnostic text data and / or a test image, and the preprocessing includes data cleaning, normalization, key feature extraction and image enhancement processing; use a natural language processing (NLP) algorithm based on a knowledge graph to match and analyze the alternative medical feature parameters corresponding to the historical test data with the first medical feature parameters to determine the similarity between the alternative medical feature parameters and the corresponding first medical feature parameters, wherein the historical test data includes existing medical test data included in a medical database serving as a target database, and the alternative medical feature parameters include the first diagnostic text data and / or the first test image corresponding to the historical test data; select alternative medical feature parameters whose similarity is greater than a preset threshold, and use the selected alternative medical feature parameters as intended medical feature parameters.

[0106] In some of the embodiments, the processing module 33 is also used to obtain alternative medical feature parameters as intended medical feature parameters, and update the first medical feature parameters with the corresponding alternative medical feature parameters to generate target medical information including target medical feature parameters, wherein the target medical feature parameters include target diagnostic text data and / or target detection images.

[0107] In some of the embodiments, the selection unit is also used to, when determining that the query operation is a consultation interaction, use the BERT natural language processing model to perform semantic understanding processing on the preprocessed inquiry text data to obtain semantic features corresponding to the inquiry text, wherein the preprocessing includes data cleaning and the response information includes the inquiry text data; in a preset medical database, multiple response text data are generated using semantic features and natural language generation (NLG) methods, wherein the alternative medical feature parameters include response text data, and the response text data is used to characterize the responses to the questions asked by the target object associated with the inquiry text data during the consultation interaction; based on the actual demand information determined by parsing the semantic features, the matching degree of each response text data with the actual demand information is calculated, and the matching degree is used as the corresponding similarity to screen at least one response text data as the intended medical feature parameter from the multiple response text data.

[0108] In some of the embodiments, the processing module 33 is further used to select the response text data with the greatest matching degree from the at least one screened response text data as the target medical information corresponding to the intended demand information.

[0109] In some of the embodiments, the medical information query device is also used to generate target recommendation information based on the target medical information after screening out the target medical information corresponding to the intended demand information, and display the target recommendation information in a preset display format on the interface of the query system, wherein the display format includes one of the following: text display, picture display, video image display.

[0110] This embodiment further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0111] Optionally, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0112] Optionally, in this embodiment, the processor may be configured to perform the following steps through a computer program: S1, after responding to the query operation triggered by the target object in the query system, obtain the query request information input by the target object based on the prompt project, wherein the query request information includes the intention demand information and the query category label information, and the query category label information is used to characterize one of the following query operations: case screening, medical popularization, information diagnosis and consultation interaction.

[0113] S2, based on the query category label information, at least using a natural language processing (NLP) algorithm to obtain candidate medical information associated with the intended demand information in the target database.

[0114] S3, in the preset matching rules, determine the target matching rule corresponding to each query operation, and according to the target matching rule, screen out the target medical information corresponding to the intended demand information from the candidate medical information, and use the target medical information as the query result.

[0115] In addition, in combination with the medical information query method based on prompt engineering in the above embodiments, the present application embodiment can provide a storage medium for implementation. The storage medium stores a computer program; when the computer program is executed by a processor, any medical information query method based on prompt engineering in the above embodiments is implemented.

[0116] Those skilled in the art should understand that the technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0117] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A medical information query method based on prompt engineering, characterized in that: include: After responding to the query operation triggered by the target object in the query system, the query request information input by the target object is obtained based on the prompt project, wherein the query request information includes intention demand information and query category label information, and the query category label information is used to characterize the query operation of one of the following: case screening, medical popularization, information diagnosis and consultation interaction; Based on the query category label information, at least using a natural language processing (NLP) algorithm, obtaining candidate medical information associated with the intended demand information in a target database; In the preset matching rules, the target matching rules corresponding to each of the query operations are determined, and according to the target matching rules, the target medical information corresponding to the intended demand information is screened out from the candidate medical information, and the target medical information is used as the query result.

2. The method according to claim 1, characterized in that After responding to the query operation triggered by the target object in the query system, the query request information input by the target object is obtained based on the prompting project, including: After receiving the query operation triggered by the target object, use the pre-trained large language model prompt project to perform multiple rounds of interaction with the corresponding target object, and obtain the target object's response information to the prompt project in each round of interaction; Determining the intended demand information input by the target object according to the response information; The query category label corresponding to the query operation is determined, and the query request information is generated according to the response information, the intended demand information and the query category label corresponding to the target object.

3. The method according to claim 2, characterized in that Based on the query category label information, at least using a natural language processing (NLP) algorithm, obtaining candidate medical information associated with the query requirement information in a target database includes: After determining the corresponding query operation according to the query category label information, acquiring the corresponding response information from the corresponding query request information; Utilizing the natural language processing (NLP) algorithm, alternative medical feature parameters associated with the first medical feature parameter are detected in the target database, and based on the similarity between the alternative medical feature parameters and the first medical feature parameter, intended medical feature parameters are screened from the alternative feature parameters, and the screened intended medical feature parameters are used as the candidate medical information, wherein the first medical feature parameter is generated by preprocessing the response information corresponding to the query operation, and the preprocessing includes at least one of the following: data cleaning, terminology standardization, key feature extraction, normalization, and image enhancement processing.

4. The method according to claim 3, characterized in that Detecting candidate medical feature parameters associated with the first medical feature parameter in the target database, and selecting intended medical feature parameters from the candidate feature parameters according to the similarity between the candidate medical feature parameters and the first medical feature parameter, including: When it is determined that the query operation is case screening, a BERT natural language processing model is used to perform semantic understanding processing on the first case data corresponding to the response information to obtain first symptom data; Using the BERT natural language processing model, detecting candidate symptom data associated with the first symptom data in a medical database serving as the target database, and acquiring intended symptom data from the candidate symptom data according to an overlap between the candidate symptom data and the first symptom data, wherein the overlap is used to characterize symptom similarity of the corresponding symptom data; After determining the second case data associated with the intended disease data, a collaborative filtering algorithm based on neighborhood and matrix decomposition is used to filter out target historical medical information from the historical medical information corresponding to the second case data to obtain the candidate medical information, wherein the historical medical information is used to characterize a treatment plan for the disease in the second case data, and the candidate medical information is used to characterize a preliminary recommended treatment plan for the disease in the first case data.

5. The method according to claim 4, characterized in that In the preset matching rules, a target matching rule corresponding to each of the query operations is determined, and according to the target matching rule, target medical information corresponding to the intended demand information is screened out from the candidate medical information, including: Determine a medical user profile of a first object associated with the first case data, wherein the first object is used to characterize a treatment object corresponding to the first case data, and the medical user profile includes user characteristics in multiple dimensions, and the user characteristics include one of the following: age, gender, physiological parameters, and medical history parameters; Filtering out a plurality of target historical medical information from all the target historical medical information corresponding to the candidate medical information according to weighted weights corresponding to all the user features of the first object; Using the trained treatment rate prediction model, the multiple selected target historical medical information are processed, and according to the predicted treatment rate corresponding to each target historical medical information, the target historical medical information with the largest predicted treatment rate is selected to obtain the target medical information, wherein the treatment rate prediction model is a prediction model trained based on one of the random forest algorithm and the extreme gradient decision tree.

6. The method according to claim 3, characterized in that Detecting candidate medical feature parameters associated with the first medical feature parameter in the target database, and selecting intended medical feature parameters from the candidate feature parameters according to the similarity between the candidate medical feature parameters and the first medical feature parameter, including: In the case where it is determined that the query operation is medical science popularization, searching for a target data entry corresponding to the medical keyword corresponding to the response information in a preset word list corresponding to a medical database as the target database, wherein the medical database stores a plurality of medical science popularization text data, and the medical science popularization text data includes at least one medical data entry; Using a preset target NLP method, the target data entry is detected in the medical data entry corresponding to each medical popular science text data, and the medical popular science text data including at least one target data entry is retrieved as candidate medical popular science text data, and the intended medical feature parameters including the candidate medical popular science text data are obtained, wherein the target NLP method includes one of the following: forward maximum matching method, reverse maximum matching method, bidirectional scanning method and word-by-word traversal method, and the similarity is determined according to the ratio of the number of the target data entry and the medical data entry in the medical popular science text data; and / or, Determining a target matching rule corresponding to each of the query operations, and screening out target medical information corresponding to the intended demand information from the candidate medical information according to the target matching rule, including: Detecting non-core items among all the medical data items corresponding to the candidate medical popular science text data; After deleting the candidate medical popular science text data of the non-core items detected, the remaining candidate medical popular science text data is used as the target medical popular science text data corresponding to the medical keywords, and the target medical popular science text data is used as the target medical information corresponding to the intended demand information.

7. The method according to claim 3, characterized in that Detecting candidate medical feature parameters associated with the first medical feature parameter in the target database, and selecting intended medical feature parameters from the candidate feature parameters according to the similarity between the candidate medical feature parameters and the first medical feature parameter, including: When it is determined that the query operation is information diagnosis and the response information includes medical test data, the pre-processed medical test data is processed by a pre-built convolutional neural network (CNN) to generate the first medical feature parameter, wherein the first medical feature parameter includes diagnostic text data and / or a test image, and the pre-processing includes data cleaning, normalization, key feature extraction, and image enhancement processing; Using a natural language processing (NLP) algorithm based on a knowledge graph, a matching analysis is performed on the candidate medical feature parameters corresponding to the historical test data and the first medical feature parameters to determine the similarity between the candidate medical feature parameters and the corresponding first medical feature parameters, wherein the historical test data includes existing medical test data included in a medical database serving as the target database, and the candidate medical feature parameters include first diagnostic text data and / or first test images corresponding to the historical test data; selecting the candidate medical feature parameters whose similarity is greater than a preset threshold, and using the selected candidate medical feature parameters as the intended medical feature parameters; and / or, In the preset matching rules, the target matching rules corresponding to each of the query operations are determined, and according to the target matching rules, the target medical information corresponding to the intended demand information is screened out from the candidate medical information, including: obtaining the alternative medical feature parameters as the intended medical feature parameters, and updating the first medical feature parameters with the corresponding alternative medical feature parameters, generating the target medical information including target medical feature parameters, wherein the target medical feature parameters include target diagnosis text data and / or target detection images.

8. The method according to claim 3, characterized in that Detecting candidate medical feature parameters associated with the first medical feature parameter in the target database, and selecting intended medical feature parameters from the candidate feature parameters according to the similarity between the candidate medical feature parameters and the first medical feature parameter, including: In the case where it is determined that the query operation is a consulting interaction, a BERT natural language processing model is used to perform semantic understanding processing on the preprocessed query text data to obtain semantic features corresponding to the query text, wherein the preprocessing includes data cleaning, and the response information includes the query text data; In a preset medical database, a plurality of response text data are generated by using the semantic features and the natural language generation (NLG) method, wherein the candidate medical feature parameters include the response text data, and the response text data is used to represent the response to the question asked by the target object associated with the inquiry text data during the consultation interaction process; According to the actual demand information determined by parsing the semantic features, the matching degree between each of the response text data and the actual demand information is calculated, and the matching degree is used as the corresponding similarity to select at least one of the response text data as the intended medical feature parameter from the plurality of response text data; and / or In the preset matching rules, the target matching rules corresponding to each of the query operations are determined, and according to the target matching rules, the target medical information corresponding to the intended demand information is screened out from the candidate medical information, including: from at least one of the screened response text data, the response text data with the largest matching degree is selected as the target medical information corresponding to the intended demand information.

9. The method according to any one of claims 4 to 8, characterized in that: After screening out the target medical information corresponding to the intended demand information, the method further includes: generating target recommendation information based on the target medical information, and displaying the target recommendation information in a preset display format on the interface of the query system, wherein the display format includes one of the following: text display, picture display, video image display.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the medical information query method based on prompt engineering described in any one of claims 1 to 9 are implemented.

Images