Intelligent diagnosis and treatment assistant construction method and device, electronic equipment and storage medium

Abstract

The invention relates to the technical field of artificial intelligence, and discloses an intelligent diagnosis and treatment assistant construction method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring multi-modal first diagnosis and treatment guide data, and analyzing the first diagnosis and treatment guide data through a preset multi-modal data processing model to obtain first diagnosis and treatment guide text data; performing semantic analysis processing on the first diagnosis and treatment guide text data through a pre-trained large language model to obtain diagnosis and treatment guide knowledge entry data; generating corresponding diagnosis and treatment knowledge graph description data according to the diagnosis and treatment guide knowledge entry data; constructing a target intelligent diagnosis and treatment knowledge base according to the diagnosis and treatment knowledge graph description data through a retrieval enhancement generation mechanism and a distributed full-text retrieval engine; and constructing a target intelligent diagnosis and treatment assistant, receiving the diagnosis and treatment question data through the target intelligent diagnosis and treatment assistant, and returning diagnosis and treatment answer data corresponding to the diagnosis and treatment question data. According to the method, the use efficiency of the diagnosis and treatment guide and the medical service quality are improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a method, device, electronic device and storage medium for constructing an intelligent diagnosis and treatment assistant. Background Art

[0002] In the development of the medical industry, the amount of knowledge is huge and updated rapidly. Various diagnosis and treatment guidelines contain a wide range of content, involving multiple disciplines. With the emergence of new clinical research results, diagnosis and treatment guidelines are updated frequently, making it difficult for doctors to track and grasp the latest diagnosis and treatment information in a timely manner. At the same time, when doctors need to make clinical decisions quickly, it is also difficult for them to quickly find the required information from the long and voluminous diagnosis and treatment information mentioned above, which leads to delays or inappropriate treatment decisions. In addition, different doctors may have differences in professional knowledge and experience when interpreting and applying diagnosis and treatment guidelines, which in turn affects the consistency of the implementation of diagnosis and treatment guidelines and the consistency of patient treatment. Summary of the invention

[0003] The main technical problem solved by the implementation methods of the present application is that the traditional way for doctors to use diagnosis and treatment guidelines is time-consuming and inefficient.

[0004] In order to solve the above-mentioned technical problems, the first technical solution adopted in the implementation mode of the present application is: to provide a method for constructing an intelligent diagnosis and treatment assistant, including: obtaining multimodal first diagnosis and treatment guideline data, parsing the first diagnosis and treatment guideline data through a preset multimodal data processing model, and obtaining first diagnosis and treatment guideline text data; performing semantic parsing processing on the first diagnosis and treatment guideline text data through a pre-trained large language model to obtain diagnosis and treatment guideline knowledge item data; generating corresponding diagnosis and treatment knowledge graph description data according to the diagnosis and treatment guideline knowledge item data; constructing a target intelligent diagnosis and treatment knowledge base according to the diagnosis and treatment knowledge graph description data through a retrieval enhancement generation mechanism and a distributed full-text retrieval engine; constructing a target intelligent diagnosis and treatment assistant, and receiving diagnosis and treatment question data through the target intelligent diagnosis and treatment assistant, and returning diagnosis and treatment answer data corresponding to the diagnosis and treatment question data, wherein the diagnosis and treatment answer data is generated according to the data in the target intelligent diagnosis and treatment knowledge base.

[0005] Optionally, the step of parsing the first diagnosis and treatment guideline data through a preset multimodal data processing model to obtain first diagnosis and treatment guideline text data includes: identifying the data modality of the first diagnosis and treatment guideline data through the multimodal data processing model; calling different diagnosis and treatment guideline data processing modules in the multimodal data processing model to perform data feature extraction processing according to the data modality to obtain multimodal diagnosis and treatment guideline data features; performing cross-modal feature data fusion processing on the multimodal diagnosis and treatment guideline data features to obtain diagnosis and treatment guideline fusion feature data; and converting the diagnosis and treatment guideline fusion feature data into the first diagnosis and treatment guideline text data.

[0006] Optionally, the step of semantically parsing the first diagnosis and treatment guideline text data by the pre-trained large language model to obtain diagnosis and treatment guideline knowledge entry data includes: converting the first diagnosis and treatment guideline text data into a diagnosis and treatment guideline text encoding sequence in the preset encoding format of the large language model; reasoning the diagnosis and treatment guideline text encoding sequence through the semantic analysis module of the large language model to mine the potential information data in the first diagnosis and treatment guideline text data to obtain a diagnosis and treatment guideline semantic parsing result; and extracting corresponding knowledge entry data from the diagnosis and treatment guideline semantic parsing result according to different preset diagnosis and treatment knowledge type templates to obtain the corresponding diagnosis and treatment guideline knowledge entry data.

[0007] Optionally, the step of generating corresponding diagnosis and treatment knowledge graph description data according to the diagnosis and treatment guideline knowledge entry data includes: parsing the diagnosis and treatment guideline knowledge entry data to obtain diagnosis and treatment guideline entities, the entity logical relationships between the diagnosis and treatment guideline entities, and the attribute data of the diagnosis and treatment guideline entities and the entity logical relationships; generating corresponding diagnosis and treatment knowledge graph nodes according to the diagnosis and treatment guideline entities, the entity logical relationships and the attribute data, and diagnosis and treatment knowledge graph edge data between the diagnosis and treatment knowledge graph nodes; and constructing the diagnosis and treatment knowledge graph description data through the diagnosis and treatment knowledge graph nodes and the diagnosis and treatment knowledge graph edge data.

[0008] Optionally, the step of constructing a target intelligent diagnosis and treatment knowledge base according to the diagnosis and treatment knowledge graph description data through a retrieval enhancement generation mechanism and a distributed full-text retrieval engine includes: importing the preprocessed diagnosis and treatment knowledge graph description data into a preset distributed full-text retrieval engine, and creating a corresponding diagnosis and treatment guideline data index during the import process; generating a diagnosis and treatment guideline interaction interface between the preset retrieval enhancement generation mechanism and the distributed full-text retrieval engine after importing the data; and constructing the target intelligent diagnosis and treatment knowledge base according to the distributed full-text retrieval engine after importing the data, the diagnosis and treatment guideline data index, and the diagnosis and treatment guideline interaction interface.

[0009] Optionally, the steps of receiving diagnostic question data through the target intelligent diagnosis and treatment assistant and returning diagnostic answer data corresponding to the diagnostic question data include: when receiving diagnostic question data sent by a user, performing preprocessing on the diagnostic question data; sending the preprocessed diagnostic question data to the diagnostic guideline interaction interface, and obtaining first diagnostic knowledge data from the target intelligent diagnosis and treatment knowledge base through the diagnostic guideline interaction interface; screening and sorting the first diagnostic knowledge data to obtain second diagnostic knowledge data with the highest relevance to the diagnostic question data; performing semantic enhancement processing on the second diagnostic knowledge data through the pre-trained large language model to obtain the diagnostic answer data; and returning the diagnostic answer data through the target intelligent diagnosis and treatment assistant.

[0010] Optionally, after the steps of constructing the target intelligent diagnosis and treatment assistant, receiving diagnostic question data through the target intelligent diagnosis and treatment assistant, and returning diagnostic answer data corresponding to the diagnostic question data, the method further includes: testing the target intelligent diagnosis and treatment assistant with a preset test sample data set to obtain a first test result; and optimizing the multi-modal data processing model, the pre-trained large language model, and the target intelligent diagnosis and treatment knowledge base according to the test result.

[0011] To solve the above technical problems, the second technical solution adopted in the embodiments of the present application is: to provide an intelligent diagnosis and treatment assistant construction device, including: a diagnostic data acquisition module, configured to acquire multi-modal first diagnostic guideline data, and parse the first diagnostic guideline data through a preset multi-modal data processing model to obtain first diagnostic guideline text data; a diagnostic data processing module, configured to perform semantic parsing processing on the first diagnostic guideline text data through a pre-trained large language model to obtain diagnostic guideline knowledge entry data; a diagnostic knowledge graph module, configured to generate corresponding diagnostic knowledge graph description data according to the diagnostic guideline knowledge entry data; a diagnostic knowledge base construction module, configured to construct a target intelligent diagnosis and treatment knowledge base according to the diagnostic knowledge graph description data through a retrieval enhancement generation mechanism and a distributed full-text retrieval engine; and an intelligent diagnosis and treatment assistant construction module, configured to construct a target intelligent diagnosis and treatment assistant, and receive diagnostic question data through the target intelligent diagnosis and treatment assistant and return diagnostic answer data corresponding to the diagnostic question data, wherein the diagnostic answer data is generated according to the data in the target intelligent diagnosis and treatment knowledge base.

[0012] To solve the above technical problems, the third technical solution adopted in the embodiments of the present application is: to provide an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the intelligent diagnosis and treatment assistant construction method as described above.

[0013] To solve the above technical problems, the fourth technical solution adopted in the embodiments of the present application is: to provide a non-volatile computer-readable storage medium, which stores computer-executable instructions, and when the computer-executable instructions are executed by an electronic device, the electronic device executes the intelligent diagnosis and treatment assistant construction method as described above.

[0014] Different from the related art, in the knowledge processing of the present application, the multi-modal data processing model integrates various forms of medical information, breaks through the limitations of a single modality, and the pre-trained large language model mines potential information in the text and extracts structured knowledge entries to lay a solid foundation for subsequent applications. At the level of knowledge presentation and management, the diagnosis and treatment knowledge graph describes the data and intuitively shows the knowledge associations in graphics, making the complex medical knowledge system clear and easy to understand, facilitating the acquisition and dissemination of knowledge. The constructed target intelligent diagnosis and treatment knowledge base has powerful functions. With the help of the retrieval-enhanced generation mechanism and the distributed full-text retrieval engine, it can efficiently store and accurately retrieve a large amount of knowledge, providing users with adapted medical knowledge. The target intelligent diagnosis and treatment assistant brings convenient one-stop services to users, can quickly give professional and accurate diagnosis and treatment suggestions, assist in scientific decision-making, and improve the accessibility of medical services. At the same time, this method can optimize the model and knowledge base according to the test feedback to make it adapt to the development of medicine and user needs, ensuring the practicality and effectiveness of the intelligent diagnosis and treatment assistant. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] One or more embodiments are exemplarily illustrated by corresponding drawings. These exemplary illustrations do not constitute limitations on the embodiments. Elements with the same reference numerals in the drawings are represented as similar elements, unless otherwise stated, and the drawings in the drawings do not constitute a proportional limitation.

[0016] Figure 1 It is a schematic diagram of the operating environment of the intelligent diagnosis and treatment assistant construction method provided by the embodiments of the present application.

[0017] Figure 2 It is a schematic diagram of the execution process of the intelligent diagnosis and treatment assistant construction method provided by the embodiments of the present application.

[0018] Figure 3 It is a schematic diagram of the execution process of obtaining the first diagnosis and treatment guideline text data in the intelligent diagnosis and treatment assistant construction method provided by the embodiments of the present application.

[0019] Figure 4 It is a schematic flowchart of constructing a target intelligent diagnosis and treatment knowledge base in the intelligent diagnosis and treatment assistant construction method provided by an embodiment of the present application.

[0020] Figure 5 It is a schematic flowchart of intelligent diagnosis and treatment in the intelligent diagnosis and treatment assistant construction method provided by an embodiment of the present application.

[0021] Figure 6 It is a schematic system structure diagram of the intelligent diagnosis and treatment assistant construction device provided by an embodiment of the present application.

[0022] Figure 7 It is a schematic hardware structure diagram of an electronic device for executing the intelligent diagnosis and treatment assistant construction method provided by an embodiment of the present application. Detailed implementation manners

[0023] In order to make the objectives, technical solutions and advantages of the present application more clear and understandable, the present application will be further described in detail below with reference to 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 used to limit the present application.

[0024] It should be noted that if there is no conflict, the various features in the embodiments of the present application can be combined with each other, and all are within the protection scope of the present application. In addition, although the functional modules are divided in the device schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different module division from that in the device schematic diagram or a different order from that in the flowchart.

[0025] Unless otherwise defined, all technical and scientific terms used in this specification have the same meaning as commonly understood by those skilled in the technical field to which this application belongs. The terms used in the specification of this application are only for the purpose of describing specific implementation manners and are not used to limit this application. The term "and / or" used in this specification includes any and all combinations of one or more of the related listed items.

[0026] For the convenience of understanding this embodiment, first, a detailed introduction to an intelligent diagnosis and treatment assistant construction method disclosed in an embodiment of the present application will be given. Please refer to Figure 1 , Figure 1 It is a schematic diagram of the operating environment of the intelligent diagnosis and treatment assistant construction method provided by an embodiment of the present application. As shown in Figure 1 , generally, the execution subject of the intelligent diagnosis and treatment assistant construction method provided by an embodiment of the present application is an electronic device with a certain computing ability, such as a computer device. In some possible implementation manners, the intelligent diagnosis and treatment assistant construction method can be implemented by a processor calling computer-readable instructions stored in a memory. Among them, Figure 1The computer device in Figure 1 can be a server. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, Content Delivery Network (CDN), and big data and artificial intelligence platforms. It can be understood that

[0027] Please continue to refer to Figure 2 , Figure 2 which is a schematic execution flow diagram of the intelligent diagnosis and treatment assistant construction method provided by the embodiments of the present application. As Figure 2 shown, it includes the following steps:

[0028] S1. Obtain multi-modal first diagnosis and treatment guideline data, and parse the first diagnosis and treatment guideline data through a preset multi-modal data processing model to obtain first diagnosis and treatment guideline text data.

[0029] Among them, with the development of precision medicine, various diagnosis and treatment guidelines are constantly updated, including a large amount of content such as molecular biology information, pathological classification, clinical staging, treatment options, and prognosis evaluation. The document formats of these contents are complex, such as PDF files, TXT files, WORD files, picture files, table files, charts, reference videos, and reference documents. Therefore, when processing various types of diagnosis and treatment guideline data, the embodiments of the present application use a multi-modal data processing model.

[0030] As an optional implementation manner, please continue to refer to Figure 3 , Figure 3 which is a schematic execution flow diagram of obtaining the first diagnosis and treatment guideline text data in the intelligent diagnosis and treatment assistant construction method provided by the embodiments of the present application. As Figure 3 shown, it includes steps S11 to S14 as follows.

[0031] S11. Identify the data modality of the first diagnosis and treatment guideline data through the multi-modal data processing model.

[0032] Among them, the data modality can be in the form of text, image, audio, etc. For example, in diabetes diagnosis and treatment guideline data, there may be several modalities as follows. Text modality: The parts that describe diabetes symptoms, diagnostic criteria, treatment methods, etc. in words, such as "The common symptoms of diabetic patients are polyuria, polydipsia, polyphagia and weight loss". Image modality: There may be fundus lesion images of diabetic patients, which are used to assist in the diagnosis of diabetic retinopathy, etc. Numerical modality: It includes various test index data such as blood glucose values and glycated hemoglobin values. For example, "A fasting blood glucose value greater than or equal to 7.0 mmol / L can be used as one of the diagnostic bases for diabetes".

[0033] S12. Call different diagnosis and treatment guideline data processing modules in the multi-modal data processing model according to the data modality to perform data feature extraction processing, and obtain multi-modal diagnosis and treatment guideline data features.

[0034] Among them, different data modalities require different processing methods to extract valuable features. Text modality processing: For the text part, the model may call the natural language processing module and extract key features through means such as lexical analysis and syntactic analysis. For example, from the sentence "Carbohydrate intake should be controlled and dietary fiber intake should be increased in diet", key feature words such as "diet control", "carbohydrate", and "dietary fiber" are extracted. Image modality processing: For diabetic fundus lesion images, the image processing module will be called, and a convolutional neural network (CNN) may be used to extract features in the image, such as the shape, size, color, etc. of the lesion area. For example, features such as abnormal dilation of microvessels and the presence of microaneurysms in the fundus image are identified. Numerical modality processing: For numerical data such as blood glucose values and glycated hemoglobin values, the processing module will calculate features such as mean, standard deviation, and change trend. For example, observing the fluctuation of blood glucose values over a period of time, if the fasting blood glucose value has been between 8.0 - 9.0 mmol / L for a consecutive week, the feature "blood glucose value is on the high side and relatively stable" is extracted.

[0035] S13. Perform cross-modal feature data fusion processing on the multi-modal diagnosis and treatment guideline data features to obtain diagnosis and treatment guideline fusion feature data.

[0036] Among them, by fusing the features extracted from different modalities, more comprehensive information can be obtained. For example, the text features indicate that the patient needs to control diet and strengthen exercise, the image features show that the patient has mild diabetic retinopathy, and the numerical features show that the blood glucose value is on the high side but relatively stable. During fusion, these features will be integrated to form a comprehensive feature description: "The patient has a high and relatively stable blood glucose value, has mild diabetic retinopathy, and needs to control diet and strengthen exercise". Through cross-modal feature fusion, more comprehensive information related to diabetes diagnosis and treatment can be understood, providing richer data support for subsequent decision-making.

[0037] S14. Convert the integrated feature data of the diagnosis and treatment guidelines into the first diagnosis and treatment guidelines text data.

[0038] Among them, through the above steps S11 to S14, from identifying the data modality, extracting multi-modal features, to cross-modal feature fusion, and finally converting it into the diagnosis and treatment guidelines text data. First, it improves the comprehensiveness of data processing. Traditional single-modal data processing cannot comprehensively cover diagnosis and treatment information. Multi-modal data processing can integrate various forms of data such as text, images, and numerical values, making the construction of diagnosis and treatment guidelines more comprehensive. Second, different data modalities reflect disease characteristics from different angles. Cross-modal feature fusion can avoid the limitations of single-modal information and more accurately grasp the disease situation, providing a basis for formulating more accurate diagnosis and treatment guidelines.

[0039] S2. Perform semantic parsing on the first diagnosis and treatment guidelines text data through a pre-trained large language model to obtain the diagnosis and treatment guidelines knowledge item data.

[0040] As an optional implementation method, the above step S2 may specifically include the following steps S21 to S23.

[0041] S21. Convert the first diagnosis and treatment guidelines text data into a diagnosis and treatment guidelines text coding sequence in the preset coding format of the large language model.

[0042] Among them, the large language model essentially processes digital data, while the original diagnosis and treatment guidelines text is in a human-readable natural language form and cannot be directly understood and processed by the model. Therefore, it is necessary to convert the text data into a coding sequence that the model can process. Different large language models use different coding algorithms. Common ones include Word Piece coding (such as the BERT model), Byte Pair Encoding (BPE, such as GPT series models), etc. These coding algorithms will split the text into individual sub-words or tokens, and then assign a unique coding number to each sub-word or token. For example, for "hypertensive patients", the coding algorithm may split it into two sub-words, "hypertension" and "patients", and then assign corresponding coding numbers to them, finally forming a coding sequence. Through coding, the natural language text is converted into a digital form, enabling the large language model to perform subsequent calculations and processing on it, laying a foundation for semantic analysis.

[0043] S22. Infer the diagnosis and treatment guidelines text coding sequence through the semantic analysis module of the large language model, mine the potential information data in the first diagnosis and treatment guidelines text data, and obtain the diagnosis and treatment guidelines semantic parsing result.

[0044] Among them, simply converting the text into an encoded sequence is not enough. It is also necessary to enable the model to understand the semantic information represented by these encodings, and to extract the implicit and less directly expressed information in the text, so as to more comprehensively and deeply understand the content of the diagnosis and treatment guidelines. The semantic analysis module of the large language model is usually obtained through large-scale pre-training, which contains rich language knowledge and semantic understanding capabilities. After inputting the encoded sequence of the diagnosis and treatment guidelines text, the module will perform a series of calculations and inferences on the sequence, and use the grammar, semantics, and context information learned by the model to identify key information such as the subject, action, condition, and suggestion in the text, and extract potential information. For example, for the sentence "Patients with hypertension should limit sodium intake", the model can not only identify that the subject is "patients with hypertension" and the action is "limit sodium intake", but may also extract potential information such as "limiting sodium intake helps control blood pressure" based on its knowledge reserve. Through semantic analysis, the encoded sequence is transformed into meaningful semantic information, providing a richer and more accurate basis for subsequent extraction of knowledge entries, and helping to deeply understand the meaning and intention of the diagnosis and treatment guidelines.

[0045] S23. Extract the corresponding knowledge entry data from the semantic analysis results of the diagnosis and treatment guidelines according to different preset diagnosis and treatment knowledge type templates to obtain the corresponding diagnosis and treatment guidelines knowledge entry data.

[0046] Among them, through the above steps S21 to S23, the diagnosis and treatment guidelines text is converted into an encoded sequence that can be processed by the model, solving the problem that the large language model is difficult to directly understand natural language, ensuring that the text data can smoothly enter the model for analysis, building a bridge for subsequent operations, and enabling the large language model to efficiently process text information in the medical field. With the powerful semantic analysis ability of the large language model, deeply excavate the potential information in the diagnosis and treatment guidelines text, break through the limitations of surface words, comprehensively and accurately understand the meaning and intention behind the text, provide richer and deeper knowledge support for medical decision-making, and help formulate more scientific and reasonable diagnosis and treatment plans. Extract knowledge entries through preset diagnosis and treatment knowledge type templates, and transform the semantic analysis results into a standardized and structured data form, greatly improving the manageability of knowledge. Structured data is convenient for storage, query, and sharing, is conducive to building a medical knowledge graph, promoting knowledge circulation and collaboration between different medical information systems, improving the overall quality and efficiency of medical services, and promoting the development of medical informatization.

[0047] S3. Generate corresponding diagnosis and treatment knowledge graph description data according to the diagnosis and treatment guidelines knowledge entry data.

[0048] As an optional implementation method, the above step S3 may specifically include the following steps S31 to S33.

[0049] S31. Analyze the data of the diagnosis and treatment guideline knowledge entries to obtain the diagnosis and treatment guideline entities, the entity logical relationships between the diagnosis and treatment guideline entities, and the attribute data of the diagnosis and treatment guideline entities and the entity logical relationships.

[0050] Among them, the structured diagnosis and treatment guideline knowledge entry data is further disassembled to extract the most critical entities, the logical connections between the entities, and the relevant attributes, preparing for constructing a knowledge graph to more clearly display the internal structure and semantic information of the diagnosis and treatment knowledge.

[0051] S32. Generate the corresponding nodes of the diagnosis and treatment knowledge graph and the data of the edges of the diagnosis and treatment knowledge graph between the nodes of the diagnosis and treatment knowledge graph according to the diagnosis and treatment guideline entities, the entity logical relationships, and the attribute data.

[0052] S33. Construct the description data of the diagnosis and treatment knowledge graph through the nodes of the diagnosis and treatment knowledge graph and the data of the edges of the diagnosis and treatment knowledge graph.

[0053] Among them, through the above steps S31 to S33, the knowledge graph integrates rich diagnosis and treatment information and logical relationships, and can provide comprehensive and accurate references for clinical decision-making. During the diagnosis and treatment process, medical staff can use the knowledge graph to analyze the possible development paths of diseases and evaluate the effectiveness of treatment plans, so as to make more scientific and reasonable decisions. The knowledge graph has good scalability and is convenient for updating and supplementing the diagnosis and treatment knowledge. With the continuous progress of medical research and the accumulation of clinical experience, new knowledge and relationships can be added to the knowledge graph in a timely manner to ensure the timeliness and integrity of the knowledge and provide continuous support for medical practice.

[0054] S4. Construct a target intelligent diagnosis and treatment knowledge base according to the description data of the diagnosis and treatment knowledge graph through a retrieval enhancement generation mechanism and a distributed full-text retrieval engine.

[0055] As an optional implementation manner, please continue to refer to Figure 4 , Figure 4 which is a schematic flowchart of constructing a target intelligent diagnosis and treatment knowledge base in the method for constructing an intelligent diagnosis and treatment assistant provided by the embodiment of the present application. As shown in Figure 4 , it includes the following steps S41 to S43.

[0056] S41. Import the preprocessed description data of the diagnosis and treatment knowledge graph into a pre-set distributed full-text retrieval engine, and create a corresponding diagnosis and treatment guideline data index during the import process. For example, the distributed full-text retrieval engine is set to Elasticsearch.

[0057] S42. Generate a diagnosis and treatment guideline interaction interface between the preset retrieval enhancement generation mechanism and the distributed full-text retrieval engine after importing the data.

[0058] S43. Construct a target intelligent diagnosis and treatment knowledge base based on the distributed full-text retrieval engine, diagnosis and treatment guideline data index, and diagnosis and treatment guideline interaction interface after importing data.

[0059] Among them, in terms of retrieval efficiency, an index is established in cooperation with the distributed full-text retrieval engine to achieve efficient and accurate search for a large amount of diagnosis and treatment knowledge, significantly shortening the time for medical staff to obtain information and avoiding information errors and omissions. In terms of interactive collaboration, the generated interaction interface enables the retrieval enhanced generation mechanism and the engine to cooperate flexibly, optimizes the retrieval results based on user characteristics, provides more demand-tailored knowledge recommendations, and realizes a more intelligent and user-friendly interaction experience. From the perspective of the overall medical support system, the constructed intelligent diagnosis and treatment knowledge base integrates knowledge management and decision support functions, can quickly provide diagnosis and treatment references according to the actual situation of patients, assist in making scientific decisions, and effectively improve the quality and efficiency of medical services.

[0060] As another alternative implementation, please continue to refer to Figure 5 , Figure 5 is a schematic flowchart of intelligent diagnosis and treatment in the intelligent diagnosis and treatment assistant construction method provided by the embodiment of the present application. As shown in Figure 5 , it includes the following steps S44 to S48.

[0061] S44. When receiving the diagnosis and treatment question data sent by the user, perform preprocessing on the diagnosis and treatment question data.

[0062] S45. Send the preprocessed diagnosis and treatment question data to the diagnosis and treatment guideline interaction interface, and obtain the first diagnosis and treatment knowledge data from the target intelligent diagnosis and treatment knowledge base through the diagnosis and treatment guideline interaction interface.

[0063] S46. Screen and sort the first diagnosis and treatment knowledge data to obtain the second diagnosis and treatment knowledge data with the highest correlation degree with the diagnosis and treatment question data.

[0064] S47. Perform semantic enhancement processing on the second diagnosis and treatment knowledge data through a pre-trained large language model to obtain diagnosis and treatment answer data.

[0065] S48. Return the diagnosis and treatment answer data through the target intelligent diagnosis and treatment assistant.

[0066] Among them, through the above steps S44 to S48, the user's diagnosis and treatment question data is preprocessed, which can make the question statement more in line with the system retrieval requirements and improve the accuracy of retrieval matching. With the help of the diagnosis and treatment guide interactive interface to obtain knowledge from the target intelligent diagnosis and treatment knowledge base, combined with the existing efficient indexing and distributed retrieval mechanism, it can quickly and accurately locate the diagnosis and treatment knowledge related to the question, save search time, and provide timely response for users. The first diagnosis and treatment knowledge data obtained is screened and sorted, and the second diagnosis and treatment knowledge data with the highest correlation with the diagnosis and treatment question can be extracted from a large amount of relevant information, helping users to quickly grasp the key content, avoid being disturbed by too much irrelevant information, and improve the efficiency of obtaining effective knowledge. The semantic enhancement processing of the second diagnosis and treatment knowledge data using the pre-trained large language model can make the expression of the diagnosis and treatment answer data more accurate, clear, and comprehensive, in line with natural language habits, and easy for users to understand. This not only improves the professionalism of the answer, but also enhances the user's experience of interacting with the intelligent diagnosis and treatment assistant. The optimized diagnosis and treatment answer data is returned by the target intelligent diagnosis and treatment assistant, providing users with a convenient and efficient channel for obtaining medical knowledge. Users do not need to search through massive amounts of medical information on their own. Smart assistants can directly provide accurate and easy-to-understand answers, meet users' knowledge needs in diagnosis and treatment, and help users make more reasonable medical decisions.

[0067] S5, constructing a target intelligent diagnosis and treatment assistant, and receiving diagnosis and treatment question data through the target intelligent diagnosis and treatment assistant, and returning diagnosis and treatment answer data corresponding to the diagnosis and treatment question data. The diagnosis and treatment answer data is generated according to the data in the target intelligent diagnosis and treatment knowledge base.

[0068] As an optional implementation, after the above step S5, the target intelligent diagnosis and treatment assistant can also be tested by a preset test sample data set to obtain a first test result. Then, the multimodal data processing model, the pre-trained large language model and the target intelligent diagnosis and treatment knowledge base are optimized according to the test results. The test sample data set provides a variety of test scenarios for the target intelligent diagnosis and treatment assistant, which can fully simulate various types of diagnosis and treatment questions in the real world. By testing these samples, the errors, deviations or incompleteness of the intelligent diagnosis and treatment assistant when answering questions can be accurately found, so as to clarify the direction that needs to be improved. The multimodal data processing model is responsible for identifying and processing different types of diagnosis and treatment data, such as text, images, numerical values, etc. The test result feedback can help discover the deficiencies of the model in data modality recognition, feature extraction, etc., and then optimize it in a targeted manner, so that it can more accurately process and integrate multi-source data, and improve the efficiency and quality of multimodal data processing. The target intelligent diagnosis and treatment knowledge base is the core knowledge source of the intelligent diagnosis and treatment assistant. Problems with missing, outdated or inaccurate knowledge in the knowledge base discovered during the testing process can be updated and supplemented in a timely manner through optimization to ensure that the content of the knowledge base is always kept up to date and most accurate.

[0069] The intelligent diagnosis and treatment assistant construction method provided by the embodiments of the present application can integrate medical information in different forms such as text, images, and numerical values by parsing the multimodal first diagnosis and treatment guideline data through a multimodal data processing model, avoiding the limitations of single-modal data, comprehensively covering diagnosis and treatment knowledge, and providing a rich and complete data basis for subsequent analysis. At the same time, the application of this model can efficiently process various types of data and improve the knowledge processing efficiency. Generate diagnosis and treatment knowledge graph description data according to the diagnosis and treatment guideline knowledge item data, and display the associations between diagnosis and treatment knowledge in an intuitive graphical structure, clarifying the complex medical knowledge system. This helps medical personnel quickly understand the logical relationships between knowledge, improve the efficiency of knowledge acquisition and dissemination, and also provides a clear knowledge framework for intelligent diagnosis and treatment. Build a target intelligent diagnosis and treatment knowledge base through a retrieval-enhanced generation mechanism and a distributed full-text retrieval engine, realizing the efficient storage and rapid retrieval of a large amount of medical knowledge. This knowledge base can not only accurately match the diagnosis and treatment questions of users, but also optimize the retrieval results according to the retrieval enhancement mechanism, providing more accurate and comprehensive knowledge support to meet the medical knowledge needs in different scenarios. The constructed target intelligent diagnosis and treatment assistant can receive diagnosis and treatment question data and return corresponding diagnosis and treatment answer data. Based on the target intelligent diagnosis and treatment knowledge base, it provides users with a convenient one-stop medical knowledge query service. Users can quickly obtain professional and accurate diagnosis and treatment suggestions without having to search through a large number of medical materials by themselves, assisting users in making scientific medical decisions and improving the accessibility and quality of medical services.

[0070] Please continue to refer to Figure 6 , Figure 6 which is a schematic system structure diagram of the intelligent diagnosis and treatment assistant construction device provided by the embodiments of the present application. As Figure 6 shown, the intelligent diagnosis and treatment assistant construction device 60 includes: a diagnosis and treatment data acquisition module 61, a diagnosis and treatment data processing module 62, a diagnosis and treatment knowledge graph module 63, a diagnosis and treatment knowledge base construction module 64, and an intelligent diagnosis and treatment assistant construction module 65.

[0071] The diagnosis and treatment data acquisition module 61 is used to acquire multimodal first diagnosis and treatment guideline data, and parse the first diagnosis and treatment guideline data through a preset multimodal data processing model to obtain first diagnosis and treatment guideline text data.

[0072] The diagnosis and treatment data processing module 62 is used to perform semantic parsing processing on the first diagnosis and treatment guideline text data through a pre-trained large language model to obtain diagnosis and treatment guideline knowledge item data.

[0073] The diagnosis and treatment knowledge graph module 63 is used to generate corresponding diagnosis and treatment knowledge graph description data according to the diagnosis and treatment guideline knowledge item data.

[0074] The diagnosis and treatment knowledge base construction module 64 is used to construct a target intelligent diagnosis and treatment knowledge base according to the diagnosis and treatment knowledge graph description data through a retrieval-enhanced generation mechanism and a distributed full-text retrieval engine.

[0075] The intelligent diagnosis and treatment assistant construction module 65 is used to construct a target intelligent diagnosis and treatment assistant, and receive diagnosis and treatment question data through the target intelligent diagnosis and treatment assistant, and return diagnosis and treatment answer data corresponding to the diagnosis and treatment question data, wherein the diagnosis and treatment answer data is generated according to the data in the target intelligent diagnosis and treatment knowledge base.

[0076] As an alternative implementation, the diagnosis and treatment data acquisition module 61 is further specifically configured to identify the data modality of the first diagnosis and treatment guideline data through the multimodal data processing model; call different diagnosis and treatment guideline data processing modules in the multimodal data processing model according to the data modality to perform data feature extraction processing to obtain multimodal diagnosis and treatment guideline data features; perform cross-modal feature data fusion processing on the multimodal diagnosis and treatment guideline data features to obtain diagnosis and treatment guideline fusion feature data; convert the diagnosis and treatment guideline fusion feature data into the first diagnosis and treatment guideline text data.

[0077] As an alternative implementation, the diagnosis and treatment data processing module 62 is specifically configured to convert the first diagnosis and treatment guideline text data into a diagnosis and treatment guideline text coding sequence in the preset coding format of the large language model; infer the diagnosis and treatment guideline text coding sequence through the semantic analysis module of the large language model to mine the potential information data in the first diagnosis and treatment guideline text data to obtain a diagnosis and treatment guideline semantic parsing result; extract corresponding knowledge entry data from the diagnosis and treatment guideline semantic parsing result according to preset different diagnosis and treatment knowledge type templates to obtain the corresponding diagnosis and treatment guideline knowledge entry data.

[0078] As an alternative implementation, the diagnosis and treatment knowledge graph module 63 is specifically configured to parse the diagnosis and treatment guideline knowledge entry data to obtain diagnosis and treatment guideline entities, the entity logical relationships between the diagnosis and treatment guideline entities, and the attribute data of the diagnosis and treatment guideline entities and the entity logical relationships; generate corresponding diagnosis and treatment knowledge graph nodes according to the diagnosis and treatment guideline entities, the entity logical relationships and the attribute data, and diagnosis and treatment knowledge graph edge data between the diagnosis and treatment knowledge graph nodes; construct the diagnosis and treatment knowledge graph description data through the diagnosis and treatment knowledge graph nodes and the diagnosis and treatment knowledge graph edge data.

[0079] As an alternative implementation, the diagnosis and treatment knowledge base construction module 64 is specifically configured to import the preprocessed description data of the diagnosis and treatment knowledge graph into a preset distributed full-text retrieval engine, and create corresponding diagnosis and treatment guideline data indexes during the import process; generate a diagnosis and treatment guideline interaction interface between the preset retrieval enhancement generation mechanism and the distributed full-text retrieval engine after the data is imported; construct the target intelligent diagnosis and treatment knowledge base according to the distributed full-text retrieval engine after the data is imported, the diagnosis and treatment guideline data indexes, and the diagnosis and treatment guideline interaction interface.

[0080] As an alternative implementation, the intelligent diagnosis and treatment assistant construction module 65 is further specifically configured to, when receiving the diagnosis and treatment question data sent by the user, perform preprocessing on the diagnosis and treatment question data; send the preprocessed diagnosis and treatment question data to the diagnosis and treatment guideline interaction interface, and obtain the first diagnosis and treatment knowledge data from the target intelligent diagnosis and treatment knowledge base through the diagnosis and treatment guideline interaction interface; perform screening and sorting on the first diagnosis and treatment knowledge data to obtain the second diagnosis and treatment knowledge data with the highest relevance to the diagnosis and treatment question data; perform semantic enhancement processing on the second diagnosis and treatment knowledge data through the pre-trained large language model to obtain the diagnosis and treatment answer data; return the diagnosis and treatment answer data through the target intelligent diagnosis and treatment assistant.

[0081] As an alternative implementation, the intelligent diagnosis and treatment assistant construction module 65 is further specifically configured to test the target intelligent diagnosis and treatment assistant through a preset test sample data set to obtain a first test result; optimize the multi-modal data processing model, the pre-trained large language model, and the target intelligent diagnosis and treatment knowledge base according to the test result.

[0082] It should be noted that the above intelligent diagnosis and treatment assistant construction device can execute the intelligent diagnosis and treatment assistant construction method provided in the embodiments of the present application, and has the corresponding functional modules and beneficial effects for executing the method. For the technical details not described in detail in the embodiments of the intelligent diagnosis and treatment assistant construction device, reference can be made to the intelligent diagnosis and treatment assistant construction method provided in the embodiments of the present application.

[0083] Please continue to refer to Figure 7 , Figure 7 which is a schematic hardware structure diagram of an electronic device 700 for executing the intelligent diagnosis and treatment assistant construction method provided in the embodiments of the present application. As Figure 7 shown, the electronic device 700 includes:

[0084] One or more processors 710 and a memory 720. Figure 7 Taking one processor 710 as an example.

[0085] The processor 710 and the memory 720 can be connected through a bus or other means. Figure 7Take the bus connection as an example.

[0086] As a non-volatile computer-readable storage medium, the memory 720 can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules, such as the program instructions / modules corresponding to the intelligent diagnosis and treatment assistant construction method in the embodiments of the present application. By running the non-volatile software programs, instructions, and modules stored in the memory 720, the processor 710 executes various functional applications and data processing of the server, that is, implements the intelligent diagnosis and treatment assistant construction method in the above method embodiments.

[0087] The memory 720 may include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created according to the use of the intelligent diagnosis and treatment assistant construction device. In addition, the memory 720 may include a high-speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage devices. In some embodiments, the memory 720 may optionally include a memory remotely set relative to the processor 710, and these remote memories can be connected to the intelligent diagnosis and treatment assistant construction device through a network. Examples of the above network include but are not limited to the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0088] The one or more modules are stored in the memory 720, and when executed by the one or more processors 710, execute the intelligent diagnosis and treatment assistant construction method in any of the above method embodiments. For example, execute the Figure 2 method steps S1 to S5 described above, Figure 3 method steps S11 to S14 described above, Figure 4 method steps S41 to S43 described above, Figure 5 method steps S44 to S48 described above, and implement Figure 6 the functions of the modules 61-65 described above.

[0089] The above product can execute the method provided by the embodiments of the present application, and has corresponding functional modules and beneficial effects for executing the method. For technical details not described in detail in this embodiment, reference can be made to the method provided by the embodiments of the present application.

[0090] The embodiments of the present application provide a non-volatile computer-readable storage medium, and the non-volatile computer-readable storage medium stores computer-executable instructions, and these computer-executable instructions are executed by one or more processors, such as Figure 7 one of the processors 710 described above, so that the one or more processors can execute the intelligent diagnosis and treatment assistant construction method in any of the above method embodiments. For example, execute theFigure 2 The method steps S1 to S5 in Figure 3 The method steps S11 to S14 in Figure 4 The method steps S41 to S43 in Figure 5 The method steps S44 to S48 in Figure 6 realize the functions of modules 61 - 65 in

[0091] An embodiment of the present application provides a computer program product, the computer program product includes a computer program stored on a non - volatile computer - readable storage medium, the computer program includes program instructions, when the program instructions are executed by the electronic device, the electronic device can execute the intelligent diagnosis and treatment assistant construction method in any of the above - mentioned method embodiments. For example, execute the Figure 2 The method steps S1 to S5 in Figure 3 The method steps S11 to S14 in Figure 4 The method steps S41 to S43 in Figure 5 The method steps S44 to S48 in Figure 6 realize the functions of modules 61 - 65 in

[0092] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0093] Through the description of the above - mentioned implementation manners, those of ordinary skill in the art can clearly understand that each implementation manner can be realized by means of software plus a general - purpose hardware platform, and of course, it can also be realized by hardware. Those of ordinary skill in the art can understand that all or part of the processes of implementing the above - mentioned embodiment methods can be completed by a computer program instructing relevant hardware. The program can be stored in a computer - readable storage medium. When the program is executed, it can include the processes of the above - mentioned method embodiments. Among them, the storage medium can be a magnetic disk, an optical disk, a read - only memory (ROM), or a random access memory (RAM), etc.

[0094] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; under the idea of the present application, the technical features in the above embodiments or different embodiments can also be combined, and the steps can be implemented in any order, and there are many other variations in different aspects of the present application as described above. For the sake of brevity, they are not provided in detail; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for constructing an intelligent diagnosis and treatment assistant, characterized in that: include: Acquire multimodal first diagnosis and treatment guideline data, and parse the first diagnosis and treatment guideline data using a preset multimodal data processing model to obtain first diagnosis and treatment guideline text data; Performing semantic parsing on the first diagnosis and treatment guideline text data through a pre-trained large language model to obtain diagnosis and treatment guideline knowledge item data; Generate corresponding diagnosis and treatment knowledge graph description data according to the diagnosis and treatment guideline knowledge entry data; Through the search enhancement generation mechanism and the distributed full-text search engine, a target intelligent diagnosis and treatment knowledge base is constructed according to the diagnosis and treatment knowledge graph description data; Construct a target intelligent diagnosis and treatment assistant, and receive diagnosis and treatment question data through the target intelligent diagnosis and treatment assistant, and return diagnosis and treatment answer data corresponding to the diagnosis and treatment question data, wherein the diagnosis and treatment answer data is generated according to the data in the target intelligent diagnosis and treatment knowledge base.

2. The method for constructing an intelligent diagnosis and treatment assistant according to claim 1, characterized in that: The step of parsing the first diagnosis and treatment guideline data by using a preset multimodal data processing model to obtain the first diagnosis and treatment guideline text data includes: Identifying the data modality of the first diagnosis and treatment guideline data by using the multimodal data processing model; Calling different diagnosis and treatment guideline data processing modules in the multimodal data processing model according to the data modality to perform data feature extraction processing to obtain multimodal diagnosis and treatment guideline data features; Performing cross-modal feature data fusion processing on the multimodal diagnosis and treatment guideline data features to obtain diagnosis and treatment guideline fusion feature data; The diagnosis and treatment guideline fusion feature data is converted into the first diagnosis and treatment guideline text data.

3. The method for constructing an intelligent diagnosis and treatment assistant according to claim 1, characterized in that: The step of performing semantic parsing processing on the first diagnosis and treatment guideline text data by using a pre-trained large language model to obtain diagnosis and treatment guideline knowledge item data includes: Converting the first diagnosis and treatment guide text data into a diagnosis and treatment guide text encoding sequence in a preset encoding format of the large language model; Inferring the diagnosis and treatment guide text encoding sequence through the semantic analysis module of the large language model, mining the potential information data in the first diagnosis and treatment guide text data, and obtaining the diagnosis and treatment guide semantic analysis result; According to preset templates of different diagnosis and treatment knowledge types, corresponding knowledge item data is extracted from the semantic parsing results of the diagnosis and treatment guideline to obtain the corresponding diagnosis and treatment guideline knowledge item data.

4. The method for constructing an intelligent diagnosis and treatment assistant according to claim 1, characterized in that: The step of generating corresponding diagnosis and treatment knowledge graph description data according to the diagnosis and treatment guideline knowledge item data includes: Parsing the diagnosis and treatment guideline knowledge item data to obtain diagnosis and treatment guideline entities, entity logical relationships between the diagnosis and treatment guideline entities, and attribute data of the diagnosis and treatment guideline entities and the entity logical relationships; Generate corresponding diagnosis and treatment knowledge graph nodes and diagnosis and treatment knowledge graph edge data between the diagnosis and treatment knowledge graph nodes according to the diagnosis and treatment guideline entity, the entity logical relationship and the attribute data; The diagnosis and treatment knowledge graph description data is constructed through the diagnosis and treatment knowledge graph nodes and the diagnosis and treatment knowledge graph edge data.

5. The method for constructing an intelligent diagnosis and treatment assistant according to claim 1, characterized in that: The step of constructing a target intelligent diagnosis and treatment knowledge base according to the diagnosis and treatment knowledge graph description data through the retrieval enhancement generation mechanism and the distributed full-text retrieval engine includes: Importing the preprocessed diagnosis and treatment knowledge graph description data into a pre-set distributed full-text search engine, and creating a corresponding diagnosis and treatment guide data index during the import process; Generate a diagnosis and treatment guide interactive interface between the preset search enhancement generation mechanism and the distributed full-text search engine after importing data; The target intelligent diagnosis and treatment knowledge base is constructed according to the distributed full-text search engine after importing data, the diagnosis and treatment guideline data index and the diagnosis and treatment guideline interactive interface.

6. The method for constructing an intelligent diagnosis and treatment assistant according to claim 5, characterized in that: The step of receiving the diagnosis and treatment question data through the target intelligent diagnosis and treatment assistant, and returning the diagnosis and treatment answer data corresponding to the diagnosis and treatment question data, includes: When receiving the diagnosis and treatment question data sent by the user, performing question data preprocessing on the diagnosis and treatment question data; Sending the preprocessed diagnosis and treatment question data to the diagnosis and treatment guide interactive interface, and acquiring the first diagnosis and treatment knowledge data from the target intelligent diagnosis and treatment knowledge base through the diagnosis and treatment guide interactive interface; Screening and sorting the first diagnosis and treatment knowledge data to obtain the second diagnosis and treatment knowledge data with the highest correlation with the diagnosis and treatment question data; Performing semantic enhancement processing on the second diagnosis and treatment knowledge data through the pre-trained large language model to obtain the diagnosis and treatment answer data; The diagnosis and treatment answer data is returned through the target intelligent diagnosis and treatment assistant.

7. The method for constructing an intelligent diagnosis and treatment assistant according to claim 1, characterized in that: After the steps of constructing the target intelligent diagnosis and treatment assistant, receiving the diagnosis and treatment question data through the target intelligent diagnosis and treatment assistant, and returning the diagnosis and treatment answer data corresponding to the diagnosis and treatment question data, the method further includes: Testing the target intelligent diagnosis and treatment assistant using a preset test sample data set to obtain a first test result; The multimodal data processing model, the pre-trained large language model and the target intelligent diagnosis and treatment knowledge base are optimized according to the test results.

8. A device for constructing an intelligent diagnosis and treatment assistant, characterized in that: include: A diagnosis and treatment data acquisition module, used to acquire multimodal first diagnosis and treatment guideline data, and parse the first diagnosis and treatment guideline data through a preset multimodal data processing model to obtain first diagnosis and treatment guideline text data; A diagnosis and treatment data processing module, used to perform semantic parsing processing on the first diagnosis and treatment guideline text data through a pre-trained large language model to obtain diagnosis and treatment guideline knowledge item data; A diagnosis and treatment knowledge graph module, used to generate corresponding diagnosis and treatment knowledge graph description data according to the diagnosis and treatment guideline knowledge entry data; A diagnosis and treatment knowledge base construction module is used to construct a target intelligent diagnosis and treatment knowledge base according to the diagnosis and treatment knowledge graph description data through a retrieval enhancement generation mechanism and a distributed full-text retrieval engine; The intelligent diagnosis and treatment assistant construction module is used to construct a target intelligent diagnosis and treatment assistant, and receive diagnosis and treatment question data through the target intelligent diagnosis and treatment assistant, and return diagnosis and treatment answer data corresponding to the diagnosis and treatment question data, wherein the diagnosis and treatment answer data is generated based on the data in the target intelligent diagnosis and treatment knowledge base.

9. An electronic device, characterized in that: include: at least one processor; as well as, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the intelligent diagnosis and treatment assistant construction method described in any one of claims 1-7.

10. A non-volatile computer-readable storage medium, characterized in that: The non-volatile computer-readable storage medium stores computer-executable instructions. When the computer-executable instructions are executed by an electronic device, the electronic device executes the method for constructing an intelligent diagnosis and treatment assistant as described in any one of claims 1 to 7.

Images