Medical record report generation method, device, equipment, storage medium and program product

Abstract

The application relates to a medical record report generation method, device, equipment, storage medium and program product, and is applied to a diagnosis device. The method comprises the following steps: firstly, acquiring question and answer data of a user in a current inquiry process; then inputting the question and answer data into a preset diagnosis model for diagnosis recognition to obtain a diagnosis result; and then generating a medical record report according to the question and answer data and the diagnosis result; the diagnosis result comprises diagnosis information and suggestion information, and the preset diagnosis model is obtained based on sample question and answer data and an initial diagnosis model. The above method trains the preset diagnosis model in advance, diagnoses the diagnosis result of the user according to the question and answer data, and generates a corresponding medical record report based on the diagnosis result and the question and answer data. This process reduces the burden of manual medical record input of doctors and improves work efficiency.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a method, apparatus, device, storage medium, and program product for generating medical record reports. Background Technology

[0002] With societal progress and improved living standards, people are paying increasing attention to their health, leading to a continuous rise in the number of patients seeking medical care. Against this backdrop, how to efficiently utilize limited medical resources to meet the growing health demands has become a significant challenge for the medical field.

[0003] Medical records, as crucial documents documenting the onset, development, and outcome of a patient's illness, as well as the process of medical treatment, play a vital role in medical care, prevention, teaching, research, and hospital management. With the development of electronic information technology, electronic medical record systems have gradually replaced traditional paper-based medical records, becoming an indispensable part of medical institutions. Currently, the process of generating electronic medical records generally involves users manually inputting large amounts of medical data into a preset medical record template, and then generating the electronic medical record based on the medical data and the preset template.

[0004] However, the above-mentioned electronic medical record generation methods suffer from low generation efficiency. Summary of the Invention

[0005] Therefore, it is necessary to provide a method, apparatus, device, storage medium, and program product for generating electronic medical record reports that can improve the efficiency of electronic medical record report generation, in response to the above-mentioned technical problems.

[0006] Firstly, this application provides a method for generating medical record reports, applied to diagnostic devices, including:

[0007] Obtain the user's question and answer data during the current consultation process;

[0008] The question-and-answer data is input into a pre-set diagnostic model for diagnostic identification, and a diagnostic result is obtained. The diagnostic result includes diagnostic information and suggestion information. The pre-set diagnostic model is trained based on the sample question-and-answer data and the initial diagnostic model.

[0009] Based on the question-and-answer data and diagnostic results, generate a medical record report.

[0010] In one embodiment, the acquisition of user question-and-answer data during the current consultation process includes:

[0011] Obtain user questions during the current consultation process;

[0012] The question data is input into the response model for identification, and the response data during the consultation process is obtained;

[0013] Question and answer data are generated based on the question and answer data.

[0014] In one embodiment, the aforementioned preset diagnostic model includes an identification model and a query model. The input of question-and-answer data into the preset diagnostic model for diagnostic identification, to obtain a diagnostic result, includes:

[0015] The question-and-answer data is input into the recognition model to identify the disease type and obtain diagnostic information.

[0016] Diagnostic information and question-and-answer data are input into the query model to retrieve medical knowledge base information and obtain suggested information.

[0017] In one embodiment, the method further includes:

[0018] Obtain standard question-and-answer data corresponding to standard medical record reports;

[0019] Standard question-and-answer data is input into the candidate diagnostic model for training to obtain the preset diagnostic model.

[0020] In one embodiment, the acquisition of standard question-and-answer data corresponding to a standard medical record report includes:

[0021] Obtain first-sample question-and-answer data from multiple first-sample users during their consultation process within a historical time period;

[0022] Each first sample question and answer data is input into the candidate diagnostic model for diagnosis and identification, resulting in multiple candidate medical record reports;

[0023] The medical record reports that meet the preset conditions are selected from multiple candidate medical record reports and used as the standard medical record reports;

[0024] The first sample question and answer data corresponding to the standard medical record report is selected from the first sample question and answer data and used as the standard question and answer data.

[0025] In one embodiment, the method further includes:

[0026] Acquire second-sample question-and-answer data from multiple second-sample users during their consultations within a historical time period; the second-sample question-and-answer data includes medical question-and-answer data and medical knowledge data.

[0027] The second sample question-and-answer data is input into the initial diagnostic model for training to obtain the candidate diagnostic model.

[0028] Secondly, this application also provides a medical record report generation apparatus, comprising:

[0029] The acquisition module is used to acquire the user's question and answer data during the current consultation process;

[0030] The identification module is used to input question-and-answer data into a preset diagnostic model for diagnostic identification and to obtain diagnostic results. The diagnostic results include diagnostic information and suggestion information. The preset diagnostic model is trained based on sample question-and-answer data and an initial diagnostic model.

[0031] The generation module is used to generate medical record reports based on question and answer data and diagnostic results.

[0032] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0033] Obtain the user's question and answer data during the current consultation process;

[0034] The question-and-answer data is input into a pre-set diagnostic model for diagnostic identification, and a diagnostic result is obtained. The diagnostic result includes diagnostic information and suggestion information. The pre-set diagnostic model is trained based on the sample question-and-answer data and the initial diagnostic model.

[0035] Based on the question-and-answer data and diagnostic results, generate a medical record report.

[0036] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0037] Obtain the user's question and answer data during the current consultation process;

[0038] The question-and-answer data is input into a pre-set diagnostic model for diagnostic identification, and a diagnostic result is obtained. The diagnostic result includes diagnostic information and suggestion information. The pre-set diagnostic model is trained based on the sample question-and-answer data and the initial diagnostic model.

[0039] Based on the question-and-answer data and diagnostic results, generate a medical record report.

[0040] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0041] Obtain the user's question and answer data during the current consultation process;

[0042] The question-and-answer data is input into a pre-set diagnostic model for diagnostic identification, and a diagnostic result is obtained. The diagnostic result includes diagnostic information and suggestion information. The pre-set diagnostic model is trained based on the sample question-and-answer data and the initial diagnostic model.

[0043] Based on the question-and-answer data and diagnostic results, generate a medical record report.

[0044] The aforementioned method, apparatus, equipment, storage medium, and program products for generating medical record reports, applied to diagnostic equipment, first acquire the user's question-and-answer data during the current consultation process, then input the question-and-answer data into a preset diagnostic model for diagnostic identification, obtaining a diagnostic result, and then generating a medical record report based on the question-and-answer data and the diagnostic result. The diagnostic result includes diagnostic information and suggestion information. The preset diagnostic model is trained based on sample question-and-answer data and an initial diagnostic model. This method, by pre-training a preset diagnostic model to diagnose the user's diagnosis based on question-and-answer data, and then generating a corresponding medical record report based on the diagnostic result and question-and-answer data, not only reduces the burden on doctors manually entering medical records and improves work efficiency, but also, by pre-training the preset diagnostic model to generate accurate diagnostic results and generating accurate and professional electronic medical records based on the diagnostic results, provides medical staff with comprehensive and detailed medical information and decision support, thereby optimizing the overall quality of medical services. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a diagram illustrating the application environment of a method for generating medical record reports in one embodiment.

[0047] Figure 2 This is a flowchart illustrating a method for generating medical record reports in one embodiment;

[0048] Figure 3 This is a flowchart illustrating the method for generating medical record reports in another embodiment;

[0049] Figure 4 This is a flowchart illustrating the method for generating medical record reports in another embodiment;

[0050] Figure 5 This is a flowchart illustrating the method for generating medical record reports in another embodiment;

[0051] Figure 6 This is a flowchart illustrating the method for generating medical record reports in another embodiment;

[0052] Figure 7 This is a flowchart illustrating the method for generating medical record reports in another embodiment;

[0053] Figure 8 This is a flowchart illustrating the method for generating medical record reports in another embodiment;

[0054] Figure 9 This is a flowchart illustrating the method for generating medical record reports in another embodiment;

[0055] Figure 10 This is a structural block diagram of a medical record report generation device in one embodiment;

[0056] Figure 11 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0058] With societal progress and improved living standards, people are paying increasing attention to their health, leading to a continuous rise in the number of patients seeking medical care. Against this backdrop, how to efficiently utilize limited medical resources to meet the growing health demands has become a significant challenge for the medical field.

[0059] Medical records, as crucial documents documenting the onset, development, and outcome of a patient's illness, as well as the process of medical treatment, play a vital role in medical care, prevention, teaching, research, and hospital management. With the development of electronic information technology, electronic medical record systems have gradually replaced traditional paper-based medical records, becoming an indispensable part of medical institutions. Electronic medical record systems record, store, and manage patient medical information through electronic devices, greatly improving the accessibility and manageability of medical information.

[0060] Currently, the process of generating electronic medical records typically involves users manually inputting large amounts of medical data into a pre-set medical record template, and then generating the electronic medical record based on the medical data and the template. This tedious filling-in method is not only inefficient but also involves repetitive work, affecting the work efficiency of medical staff. Furthermore, existing systems often cannot automatically match patients' basic information and medical record characteristics, making it difficult to provide medical staff with relevant historical medical record references. This limits the potential of electronic medical record systems in clinical decision support.

[0061] Furthermore, existing electronic medical record systems still rely on semi-automatic operations by doctors during data entry and processing. This not only distracts doctors but may also affect their ability to observe and diagnose patients carefully. Doctors need to fill out medical records while seeing patients, which can reduce work efficiency, and completing a large number of records in a short period may affect the comprehensive assessment of the patient's condition, thus impacting the accuracy of the diagnosis.

[0062] Having described the background technology of the medical record report generation method provided in the embodiments of this application, the implementation environment involved in the medical record report generation method provided in the embodiments of this application will be briefly described below. The medical record report generation method provided in the embodiments of this application can be applied to, for example... Figure 1 The implementation environment shown includes a diagnostic device 104 and a data storage system 102. The data storage system 102 can store the data that the diagnostic device 104 needs to process. The data storage system 102 can be integrated into the diagnostic device 104 or placed in the cloud or on another network server. The data storage system 102 can collect and pre-store user question-and-answer data from historical consultations. The diagnostic device 104 can retrieve the user's question-and-answer data from the data storage system 102 for the current consultation, predict the user's physical condition based on this data, obtain the user's diagnosis, and generate a medical record report based on the user's question-and-answer data and diagnosis.

[0063] In other possible implementations, the method for generating medical record reports provided in this application embodiment can also be applied to a terminal. The terminal can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can be smart speakers, smart TVs, smart in-vehicle devices, etc. Portable wearable devices can be smartwatches, smart bracelets, head-mounted devices, etc.

[0064] Having described the application scenarios of the medical record report generation method provided in the embodiments of this application above, the following focuses on the medical record report generation method described in this application.

[0065] In one embodiment, such as Figure 2 As shown, a method for generating medical record reports is provided, which can be applied to... Figure 1 Taking diagnostic device 104 as an example, the following steps are included:

[0066] S201. Obtain the user's question and answer data during the current consultation process.

[0067] Among them, question-and-answer data refers to the question data of user A and the answer data determined by user B based on the question data during the consultation process. User A can be a patient, and user B can be a doctor. It can also be a neural network trained to identify the question data and determine the corresponding answer data based on the identification results.

[0068] In this embodiment of the application, when it is necessary to generate a medical record report efficiently and accurately based on the user's question and answer data during the consultation process, the user's question and answer data during the current consultation process can be obtained first.

[0069] Optionally, a response model can be pre-trained so that when the user inputs question data into the response model, the response model outputs response data, and when the user inputs different question and answer data multiple times into the response model, the response model outputs corresponding response data according to the different question and answer data. Then, the question and answer data in the consultation process is obtained by merging multiple different question and answer data and their corresponding response data.

[0070] Optionally, during the consultation process between the user and the doctor, the user's questions and the doctor's answers can be recorded at any time to obtain multiple sets of user question data and doctor's answer data. Based on the multiple sets of question data and answer data, question-and-answer data can be generated and uploaded to the server so that the server can obtain the user's question-and-answer data during the current consultation process.

[0071] S202. Input the question and answer data into the preset diagnostic model for diagnostic identification and obtain the diagnostic results.

[0072] The diagnostic results include diagnostic information and suggested information. The diagnostic information includes details such as the location, severity, and type of the symptom. The suggested information refers to treatment recommendations provided by the pre-defined diagnostic model based on the user's diagnostic information. For example, if the diagnostic information is "leg joint sports injury," the corresponding suggested information would include "bed rest, hot compresses, and maintaining mild joint movement."

[0073] The preset diagnostic model is trained based on sample question-and-answer data and the initial diagnostic model.

[0074] In this embodiment of the application, after obtaining the user's question and answer data during the current consultation process, the question and answer data can be input into a pre-trained preset diagnostic model for diagnostic identification to obtain a diagnostic result.

[0075] The pre-defined diagnostic model includes an identification network, a prediction network, and a query network. Optionally, after obtaining the user's question-and-answer data during the current consultation process, the question-and-answer data can be first input into the identification network to identify the disease type and obtain the user's corresponding disease type. Then, the disease type and question-and-answer data can be input into the prediction network to predict the user's diagnosis information. Finally, the user's diagnosis information can be input into the query network to retrieve the corresponding suggestion information, thereby outputting the user's diagnosis information and suggestion information.

[0076] S203. Generate a medical record report based on the question and answer data and the diagnosis results.

[0077] In this embodiment of the application, after obtaining the user's question-and-answer data during the current consultation process and the diagnostic result predicted based on the question-and-answer data, a medical record report can be generated based on the user's question-and-answer data, the diagnostic result, and the medical record report template. Optionally, a medical record report template can be obtained first, and the question-and-answer data and diagnostic result can be input into the corresponding positions in the medical record report template to obtain the medical record report.

[0078] The method for generating medical record reports provided in this application embodiment is applied to diagnostic equipment. It first acquires the user's question-and-answer data during the current consultation process, then inputs the data into a preset diagnostic model for diagnostic identification to obtain a diagnostic result. Finally, based on the question-and-answer data and the diagnostic result, a medical record report is generated. The diagnostic result includes diagnostic information and suggested information. The preset diagnostic model is trained based on sample question-and-answer data and an initial diagnostic model. This method, by pre-training a preset diagnostic model to diagnose the user's diagnosis based on the question-and-answer data, and then generating a corresponding medical record report based on the diagnostic result and question-and-answer data, not only reduces the burden of manual medical record entry for doctors and improves work efficiency, but also, by pre-training the preset diagnostic model to generate accurate diagnostic results and generating accurate and professional electronic medical records based on the diagnostic results, provides medical staff with comprehensive and detailed medical information and decision support, thereby optimizing the overall quality of medical services.

[0079] In one embodiment, in Figure 2 Based on the illustrated embodiment, the method for obtaining user question-and-answer data during the consultation process can be described in detail, such as... Figure 3 As shown, the above-mentioned S201 "obtaining user's question and answer data during the current consultation process" includes:

[0080] S301. Obtain the user's question data during the current consultation process.

[0081] In this embodiment of the application, before obtaining the user's question and answer data during the current consultation, the user's question data during the current consultation can be obtained first; optionally, the user can input the corresponding question data on the diagnostic device so that the diagnostic device can obtain the user's question data during the current consultation.

[0082] S302. Input the question data into the response model for identification to obtain the response data during the consultation process.

[0083] The response model is used to generate corresponding answer data based on the user's question data. The response model can be pre-trained using labeled sample question data and an initial response model.

[0084] In this embodiment of the application, after obtaining the user's question data during the current consultation process, the question data can be input into the response model for recognition to obtain the response data during the consultation process.

[0085] Optionally, the response model includes an identification network and a query network. After obtaining the user's question data during the current consultation process, the question data can first be input into the identification network for identification to obtain the consultation type. Then, the consultation type and question data can be input into the query network for identification to obtain the response data corresponding to the question data.

[0086] It should be noted that the response model can not only determine the corresponding response data based on the user's question data, but also guide the user to ask questions to ensure the completeness and accuracy of the question and answer data obtained in the final consultation process.

[0087] S303. Generate question and answer data based on the question data and the answer data.

[0088] In this embodiment of the application, after obtaining the question data and answer data as described above, the question data and answer data are combined to generate corresponding question-and-answer data. Optionally, the user can sequentially input multiple question data on the diagnostic device. The diagnostic device identifies multiple answer data based on the user's multiple question data, and after the user stops asking questions, it combines each question data and its corresponding answer data to obtain multiple sets of question-and-answer data.

[0089] The question-and-answer data generation method provided in this application embodiment realizes the response to question data through a pre-trained response model, and obtains question-and-answer data by combining question data and response data, thus providing a certain data foundation for the subsequent generation of medical record reports based on question-and-answer data.

[0090] In one embodiment, in Figure 2 Based on the illustrated embodiment, the aforementioned preset diagnostic model includes an identification model and a query model, and the process of obtaining diagnostic results can be described in detail, such as... Figure 4 As shown, the above-mentioned S202 "inputting question-and-answer data into a preset diagnostic model for diagnostic identification and obtaining diagnostic results" includes:

[0091] S401. Input the question-and-answer data into the recognition model to identify the disease type and obtain diagnostic information.

[0092] In this embodiment of the application, after obtaining the question and answer data as described above, the question and answer data can be input into the recognition model for recognition to obtain the user's diagnostic information.

[0093] S402. Input the diagnostic information and question-and-answer data into the query model to retrieve medical knowledge base information and obtain suggested information.

[0094] In this embodiment of the application, after obtaining the user's diagnostic information and the user's question and answer data during the diagnostic process, the diagnostic information and question and answer data are input into the query model for medical database retrieval to obtain the suggested information corresponding to the diagnostic information.

[0095] The diagnostic result acquisition method provided in this application embodiment achieves diagnostic identification and obtains diagnostic results based on question and answer data through the joint use of an identification model and a query model, providing a certain data foundation for the subsequent generation of corresponding medical record reports based on diagnostic results and question and answer data.

[0096] In one embodiment, in Figures 2-4 Based on any of the illustrated embodiments, such as Figure 5 As shown, the above method also includes:

[0097] S204. Obtain the standard Q&A data corresponding to the standard medical record report.

[0098] A standard medical record report refers to a medical record that meets preset conditions, which may be conditions that have been verified by medical personnel and meet industry standards. The standard question-and-answer data corresponding to a standard medical record report refers to the question-and-answer data used when generating the standard medical record report.

[0099] It should be noted that the medical record reports corresponding to each question and answer can be pre-stored in the diagnostic device's storage database.

[0100] In this embodiment of the application, multiple medical record reports can be obtained from the storage database first, and a standard medical record report can be determined from the multiple medical record reports according to preset conditions. After the standard medical record report is determined, the question and answer data corresponding to the standard medical record report can be retrieved from the storage database, and the question and answer data corresponding to the standard medical record report can be determined as standard question and answer data.

[0101] Optionally, the following provides a method for obtaining standard question-and-answer data corresponding to a standard medical record report. See [link to relevant documentation]. Figure 6 The aforementioned S204, "Obtaining standard question-and-answer data corresponding to standard medical record reports," includes:

[0102] S501. Obtain first sample question and answer data of multiple first sample users during the consultation process within a historical time period.

[0103] The first sample user is defined to distinguish it from the second sample user mentioned below; in essence, it comprises multiple users. The first sample question-and-answer data is the question-and-answer data corresponding to the first sample user. The first sample question-and-answer data is also defined to distinguish it from the second sample question-and-answer data mentioned below; in essence, it is also question-and-answer data.

[0104] The first sample question and answer data aims to guide first sample users to provide more detailed information about their condition. The questions in the first sample question and answer data cover information such as symptom description, past medical history, lifestyle habits, and family medical history to ensure that the medical information of first sample users can be collected comprehensively.

[0105] In this embodiment of the application, before obtaining the medical record report, it is also necessary to obtain the first sample question and answer data of multiple first sample users during the consultation process in a historical time period.

[0106] S502. Input the question-and-answer data of each first sample into the candidate diagnostic model for diagnosis and identification, and obtain multiple candidate medical record reports.

[0107] The candidate medical record reports include standard medical record reports that meet the preset conditions, as well as ordinary medical record reports that do not meet the preset conditions.

[0108] In this embodiment of the application, after obtaining multiple first sample question-and-answer data, each first sample question-and-answer data can be input into the candidate diagnostic model for diagnosis and identification to obtain the first diagnostic result corresponding to each first sample question-and-answer data. Then, based on each first sample question-and-answer data and the first diagnostic result corresponding to each first sample question-and-answer data, multiple candidate medical record reports are generated.

[0109] Optionally, a set of rules for generating medical record reports can be pre-defined. After obtaining the first sample question-and-answer data and the first diagnostic result, corresponding candidate medical record reports can be generated based on the rules. It should be noted that the rules for generating medical record reports should include the sorting and formatting of information, as well as the use of necessary medical terminology.

[0110] S503. Select medical record reports that meet the preset conditions from multiple candidate medical record reports and use them as standard medical record reports.

[0111] In this embodiment of the application, after obtaining multiple candidate medical record reports, it is determined whether each candidate medical record report meets the preset conditions. If the candidate medical record report meets the preset conditions, it is determined as a standard medical record report. If the candidate medical record report does not meet the preset conditions, it is determined as an ordinary medical record report.

[0112] Optionally, the generated candidate medical record reports can be manually scored to obtain a score for each candidate report. Candidate medical record reports that meet the preset score threshold can be used as standard medical record reports. For example, medical professionals can be organized to conduct quality assessments on multiple candidate medical record reports. The assessment criteria should include the completeness, accuracy, logic, and compliance with medical document standards. At the same time, a detailed scoring system should be established to quantitatively assess the quality of each candidate medical record report. The scoring system should cover all key parts of the medical record, such as the patient's chief complaint, present illness, past medical history, physical examination, auxiliary examinations, diagnosis, and treatment plan.

[0113] S504. Select the first sample question and answer data that corresponds to the standard medical record report from the first sample question and answer data, and use it as the standard question and answer data.

[0114] In this embodiment of the application, after determining the standard medical record report from multiple candidate medical record reports, the first sample question and answer data corresponding to the standard medical record report can be selected from the first sample question and answer data as the standard question and answer data.

[0115] This provides a method for obtaining standard question-and-answer data corresponding to standard medical record reports. By screening standard medical record reports from multiple candidate reports and training candidate diagnostic models based on the standard question-and-answer data corresponding to the standard medical record reports, a preset diagnostic model can be obtained. This allows for timely understanding of the strengths and weaknesses of candidate diagnostic models in generating medical record reports, thus paying particular attention to missing information, inaccurate diagnoses, and inappropriate treatment recommendations in medical records. In addition, by determining the standard question-and-answer data corresponding to standard medical record reports, it is possible to more effectively guide candidate diagnostic models to generate more accurate and comprehensive medical record content.

[0116] S205. Input the standard question-and-answer data into the candidate diagnostic model for training to obtain the preset diagnostic model.

[0117] Among them, the candidate diagnostic model refers to the diagnostic model after the initial diagnostic model has been pre-trained. The diagnostic accuracy of the initial diagnostic model is lower than that of the preset diagnostic model.

[0118] In this embodiment of the application, after obtaining the standard question-and-answer data, the standard question-and-answer data can be input into the candidate diagnostic model to train the candidate diagnostic model and obtain a preset diagnostic model. Optionally, after obtaining the standard question-and-answer data, this standard question-and-answer data can promote deeper and more comprehensive doctor-patient dialogue, thereby generating higher-quality medical record reports.

[0119] Optionally, the selected standard medical record reports and standard question-and-answer data can be integrated to form a training dataset for the second round of fine-tuning. This dataset can then be used to fine-tune the candidate diagnostic model in a second round. This process may involve further optimizing the model's parameters to better suit the high-quality medical record generation task.

[0120] It should be noted that before the second round of fine-tuning of the candidate diagnostic model, necessary checks and adjustments need to be made to the candidate diagnostic model, including confirming the model structure, cleaning up the model parameters, and setting an appropriate learning rate.

[0121] Optionally, during the fine-tuning of the candidate diagnostic model, if problems are found in certain aspects, such as improper handling of certain types of medical record information, the training strategy or model structure should be adjusted promptly. Additionally, based on performance monitoring and evaluation results, multiple iterative fine-tuning iterations may be necessary, with each iteration optimizing based on the previous results. After completing the second round of fine-tuning, a comprehensive evaluation of the pre-set diagnostic model is conducted using a test set to ensure that the model can stably generate high-quality electronic medical records in practical applications. Once the model demonstrates satisfactory performance on the test set, it is ready to be deployed to a real-world medical environment for use by doctors and patients. After deployment, a continuous feedback mechanism should be established to collect feedback from doctors and patients to facilitate further optimization and upgrades of the pre-set diagnostic model in the future.

[0122] This application provides a training method for a pre-defined diagnostic model, which trains the diagnostic model based on standard question-and-answer data corresponding to standard medical record reports. Since the standard question-and-answer data corresponding to standard medical record reports is filtered question-and-answer data, the diagnostic model trained based on the filtered question-and-answer data is more accurate, thus providing a certain data foundation for subsequent accurate diagnostic results to be predicted based on the accurate diagnostic model, and in turn, generating accurate medical record reports.

[0123] In one embodiment, in Figure 5 Based on the illustrated embodiments, as Figure 7 As shown, the above method also includes:

[0124] S206. Obtain second sample question and answer data of multiple second sample users during the consultation process within a historical time period.

[0125] The second sample of question-and-answer data includes medical question-and-answer data and medical knowledge data. The medical knowledge data comprises three parts: a knowledge graph, a medical knowledge guide base, and a medical knowledge question bank. The medical question-and-answer data integrates online medical consultation dialogues, real doctor-patient interactions, and question-and-answer dialogues generated by a general-purpose model based on the orthopedic medical knowledge graph. The knowledge graph is built on publicly available datasets and has undergone visualization processing, providing the model with a professional and detailed medical knowledge foundation. The medical knowledge guide base compiles professional literature, encyclopedias, and medical articles written by real doctors. The medical knowledge question bank includes question banks for medical professional examinations such as the Physician Qualification Examination and the Rehabilitation Medicine Treatment Technology Qualification Examination.

[0126] The second sample users are defined to distinguish them from the first sample users mentioned above; in essence, they are multiple users. The second sample question-and-answer data consists of the question-and-answer data corresponding to the second sample users. This second sample question-and-answer data is also defined to distinguish it from the first sample question-and-answer data mentioned above; in essence, it is also question-and-answer data.

[0127] In this embodiment, before training the candidate diagnostic model, it is necessary to obtain second-sample question-and-answer data from multiple second-sample users during their consultations over historical time periods. Optionally, after obtaining the second-sample question-and-answer data, it can be cleaned and filtered, and combined with manual sampling review and intelligent evaluation to ensure the accuracy and professionalism of the data. The second-sample question-and-answer data provides a new research direction and practical foundation for the application of large language models in specific medical fields, and is expected to significantly improve the efficiency and accuracy of medical consultations, providing patients with more professional and personalized medical advice.

[0128] S207. Input the second sample question-and-answer data into the initial diagnostic model for training to obtain the candidate diagnostic model.

[0129] The initial diagnostic model can be the Bidirectional Encoder Representation from Transformers (BERT), the Generative Pre-Trained Transformer (GPT), or a variant thereof. These models have already demonstrated strong performance in the field of natural language processing.

[0130] In this embodiment, after obtaining the second sample question-and-answer data, the second sample question-and-answer data can be input into the initial diagnostic model for training to obtain a candidate diagnostic model. Optionally, after obtaining the second sample question-and-answer data, preprocessing can be performed on the second sample question-and-answer data, including text cleaning, word segmentation, stop word removal, and standardization, to ensure data quality. Furthermore, the second sample question-and-answer data needs to be labeled so that the model can identify and learn the intent, entities, and relationships within the second sample question-and-answer data.

[0131] It should be noted that during the training process of the candidate diagnostic model, structured knowledge such as knowledge graphs, medical guideline bases, and medical question banks can be integrated into the training process to enhance the candidate diagnostic model's understanding and reasoning ability regarding medical knowledge.

[0132] Optionally, after inputting the second sample question-and-answer data into the initial diagnostic model for training and obtaining the candidate diagnostic model, the second sample question-and-answer data can be manually labeled. For example, a team of professional medical personnel and data scientists can be organized to manually label the medical entities (such as diseases, symptoms, drugs), intentions (such as inquiries, suggestions), and emotional states (such as anxiety, satisfaction) in the collected second sample question-and-answer data.

[0133] Optionally, a detailed set of manual annotation specifications and guidelines can be developed in advance to ensure that all annotators can perform manual annotation in accordance with uniform standards. The manual annotation specifications and guidelines should include clear definitions of various medical entities and intentions, as well as matters that need to be noted during the annotation process.

[0134] Furthermore, after the second sample question-and-answer data is manually annotated, a rigorous quality control process can be implemented, including regular annotation quality checks and evaluations. For example, random sampling and cross-validation methods can be used to ensure the accuracy and consistency of the manual annotations. Additionally, experienced clinical physicians can review the manually annotated results to ensure the accuracy of the annotated medical information and that it meets the requirements of clinical practice.

[0135] Optionally, the labeled second sample question-and-answer data can be organized, that is, the labeled data can be organized to form a correspondence between the second sample question-and-answer data and the corresponding labeled text, and classified according to the type of question-and-answer data to obtain the correspondence between the second sample question-and-answer data and the corresponding labeled text under each type, so as to facilitate subsequent model training and analysis.

[0136] Furthermore, after manually annotating the second sample question-and-answer data, the model parameters of the candidate diagnostic model can be fine-tuned based on the annotated second sample question-and-answer data to ensure the prediction accuracy of the candidate diagnostic model. Optionally, the second sample question-and-answer data is manually annotated to obtain annotated second sample question-and-answer data. Then, the second sample question-and-answer data is input into the pre-trained candidate diagnostic model for prediction to obtain the prediction result. The annotated second sample question-and-answer data and the prediction result are then compared to obtain the comparison result. Based on the comparison result, the model parameters of the candidate diagnostic model are fine-tuned. The model parameters include the learning rate, the depth of fine-tuning (whether to fine-tune only the top layer or the entire model), and the regularization method.

[0137] The training method for the candidate diagnostic model provided in this application provides a certain data foundation for the subsequent training of the preset diagnostic model based on the candidate diagnostic model.

[0138] In one embodiment, such as Figure 8 As shown, a method for generating medical record reports is also provided, including:

[0139] S10. Obtain the user's questions during the current consultation process;

[0140] S11. Input the question data into the response model for recognition to obtain the response data during the consultation process;

[0141] S12. Generate question-and-answer data based on the question data and answer data;

[0142] S13. Input the question-and-answer data into the recognition model to identify the disease type and obtain diagnostic information;

[0143] S14. Input the diagnostic information and question-and-answer data into the query model to retrieve medical knowledge base information and obtain suggested information;

[0144] S15. Generate a medical record report based on the question and answer data and the diagnosis results.

[0145] In one embodiment, such as Figure 9 As shown, a training method for a pre-defined diagnostic model is also provided, including:

[0146] S20. Obtain first sample question and answer data of multiple first sample users during the consultation process within a historical time period;

[0147] S21. Input the question-and-answer data of each first sample into the candidate diagnostic model for diagnosis and identification, and obtain multiple candidate medical record reports;

[0148] S22. Select medical record reports that meet the preset conditions from multiple candidate medical record reports and use them as standard medical record reports;

[0149] S23. Select the first sample question and answer data that corresponds to the standard medical record report from the first sample question and answer data, and use it as the standard question and answer data;

[0150] S24. Obtain second-sample question-and-answer data of multiple second-sample users during the consultation process within a historical time period;

[0151] S25. Input the second sample question-and-answer data into the initial diagnostic model for training to obtain the candidate diagnostic model;

[0152] S26. Input the standard question-and-answer data into the candidate diagnostic model for training to obtain the preset diagnostic model.

[0153] The above method pre-trains a preset diagnostic model to diagnose the user's diagnosis based on question-and-answer data, and then generates a corresponding medical record report based on the diagnosis and question-and-answer data. This process not only reduces the burden on doctors to manually enter medical records and improves work efficiency, but also generates accurate diagnostic results by pre-training the preset diagnostic model and generates accurate and professional electronic medical records based on the diagnostic results, providing medical staff with comprehensive and detailed medical information and decision support, thereby optimizing the overall quality of medical services.

[0154] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0155] Based on the same inventive concept, this application also provides a medical record report generation apparatus for implementing the above-described method for generating medical record reports. The solution provided by this apparatus is similar to the implementation described in the above-described method. Therefore, the specific limitations of one or more embodiments of the medical record report generation apparatus provided below can be found in the limitations of the medical record report generation method described above, and will not be repeated here.

[0156] In one exemplary embodiment, such as Figure 10 As shown, a medical record report generation device is provided, comprising: an acquisition module 10, an identification module 11, and a generation module 12, wherein:

[0157] The acquisition module 10 is used to acquire the user's question and answer data during the current consultation process.

[0158] The identification module 11 is used to input question-and-answer data into a preset diagnostic model for diagnostic identification and to obtain diagnostic results. The diagnostic results include diagnostic information and suggestion information. The preset diagnostic model is trained based on sample question-and-answer data and an initial diagnostic model.

[0159] The generation module 12 is used to generate medical record reports based on question and answer data and diagnostic results.

[0160] In an exemplary embodiment, the acquisition module 10 includes: an acquisition unit, an identification unit, and a generation unit, wherein:

[0161] The acquisition unit is specifically used to acquire the user's questions during the current consultation process;

[0162] The identification unit is specifically used to input the question data into the response model for identification, and obtain the response data during the consultation process;

[0163] The generation unit is specifically used to generate question-and-answer data based on the question data and the answer data.

[0164] In an exemplary embodiment, the preset diagnostic model includes an identification model and a query model, and the identification module 11 includes: an identification unit and a retrieval unit, wherein:

[0165] The identification unit is specifically used to input question-and-answer data into the identification model to identify the disease type and obtain diagnostic information.

[0166] The retrieval unit is specifically used to input diagnostic information and question-and-answer data into the query model to retrieve medical knowledge base information and obtain suggested information.

[0167] In an exemplary embodiment, the above apparatus further includes: an acquisition module and a training module, wherein:

[0168] The acquisition module is used to acquire standard question and answer data corresponding to standard medical record reports;

[0169] The training module is used to input standard question-and-answer data into the candidate diagnostic model for training, so as to obtain the preset diagnostic model.

[0170] In an exemplary embodiment, the above-described acquisition module includes: an acquisition unit, an identification unit, a first determination unit, and a second determination unit, wherein:

[0171] The acquisition unit is specifically used to acquire first sample question-and-answer data of multiple first sample users during the consultation process within a historical time period.

[0172] The identification unit is specifically used to input the question-and-answer data of each first sample into the candidate diagnostic model for diagnosis and identification, and to obtain multiple candidate medical record reports.

[0173] The first determining unit is specifically used to select medical record reports that meet preset conditions from multiple candidate medical record reports as standard medical record reports;

[0174] The second determining unit is specifically used to select the first sample question and answer data corresponding to the standard medical record report from the first sample question and answer data, and use it as the standard question and answer data.

[0175] In an exemplary embodiment, the above apparatus further includes: an acquisition module and a training model, wherein:

[0176] The acquisition module is used to acquire second-sample question-and-answer data from multiple second-sample users during their consultations within a historical time period; the second-sample question-and-answer data includes medical question-and-answer data and medical knowledge data.

[0177] The training module is used to input the second sample question-and-answer data into the initial diagnostic model for training, so as to obtain the candidate diagnostic model.

[0178] The modules in the aforementioned medical record report generation device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0179] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 11 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores question-and-answer data during the consultation process. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a question-and-answer data method.

[0180] Those skilled in the art will understand that Figure 11 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0181] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0182] Obtain the user's question and answer data during the current consultation process;

[0183] The question-and-answer data is input into a pre-set diagnostic model for diagnostic identification, and a diagnostic result is obtained. The diagnostic result includes diagnostic information and suggestion information. The pre-set diagnostic model is trained based on the sample question-and-answer data and the initial diagnostic model.

[0184] Based on the question-and-answer data and diagnostic results, generate a medical record report.

[0185] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0186] Obtain user questions during the current consultation process;

[0187] The question data is input into the response model for identification, and the response data during the consultation process is obtained;

[0188] Question and answer data are generated based on the question and answer data.

[0189] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0190] The question-and-answer data is input into the recognition model to identify the disease type and obtain diagnostic information.

[0191] Diagnostic information and question-and-answer data are input into the query model to retrieve medical knowledge base information and obtain suggested information.

[0192] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0193] Obtain standard question-and-answer data corresponding to standard medical record reports;

[0194] Standard question-and-answer data is input into the candidate diagnostic model for training to obtain the preset diagnostic model.

[0195] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0196] Obtain first-sample question-and-answer data from multiple first-sample users during their consultation process within a historical time period;

[0197] Each first sample question and answer data is input into the candidate diagnostic model for diagnosis and identification, resulting in multiple candidate medical record reports;

[0198] The medical record reports that meet the preset conditions are selected from multiple candidate medical record reports and used as the standard medical record reports;

[0199] The first sample question and answer data corresponding to the standard medical record report is selected from the first sample question and answer data and used as the standard question and answer data.

[0200] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0201] Acquire second-sample question-and-answer data from multiple second-sample users during their consultations within a historical time period; the second-sample question-and-answer data includes medical question-and-answer data and medical knowledge data.

[0202] The second sample question-and-answer data is input into the initial diagnostic model for training to obtain the candidate diagnostic model.

[0203] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0204] Obtain the user's question and answer data during the current consultation process;

[0205] The question-and-answer data is input into a pre-set diagnostic model for diagnostic identification, and a diagnostic result is obtained. The diagnostic result includes diagnostic information and suggestion information. The pre-set diagnostic model is trained based on the sample question-and-answer data and the initial diagnostic model.

[0206] Based on the question-and-answer data and diagnostic results, generate a medical record report.

[0207] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0208] Obtain user questions during the current consultation process;

[0209] The question data is input into the response model for identification, and the response data during the consultation process is obtained;

[0210] Question and answer data are generated based on the question and answer data.

[0211] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0212] The question-and-answer data is input into the recognition model to identify the disease type and obtain diagnostic information.

[0213] Diagnostic information and question-and-answer data are input into the query model to retrieve medical knowledge base information and obtain suggested information.

[0214] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0215] Obtain standard question-and-answer data corresponding to standard medical record reports;

[0216] Standard question-and-answer data is input into the candidate diagnostic model for training to obtain the preset diagnostic model.

[0217] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0218] Obtain first-sample question-and-answer data from multiple first-sample users during their consultation process within a historical time period;

[0219] Each first sample question and answer data is input into the candidate diagnostic model for diagnosis and identification, resulting in multiple candidate medical record reports;

[0220] The medical record reports that meet the preset conditions are selected from multiple candidate medical record reports and used as the standard medical record reports;

[0221] The first sample question and answer data corresponding to the standard medical record report is selected from the first sample question and answer data and used as the standard question and answer data.

[0222] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0223] Acquire second-sample question-and-answer data from multiple second-sample users during their consultations within a historical time period; the second-sample question-and-answer data includes medical question-and-answer data and medical knowledge data.

[0224] The second sample question-and-answer data is input into the initial diagnostic model for training to obtain the candidate diagnostic model.

[0225] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps:

[0226] Obtain the user's question and answer data during the current consultation process;

[0227] The question-and-answer data is input into a pre-set diagnostic model for diagnostic identification, and a diagnostic result is obtained. The diagnostic result includes diagnostic information and suggestion information. The pre-set diagnostic model is trained based on the sample question-and-answer data and the initial diagnostic model.

[0228] Based on the question-and-answer data and diagnostic results, generate a medical record report.

[0229] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0230] Obtain user questions during the current consultation process;

[0231] The question data is input into the response model for identification, and the response data during the consultation process is obtained;

[0232] Question and answer data are generated based on the question and answer data.

[0233] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0234] The question-and-answer data is input into the recognition model to identify the disease type and obtain diagnostic information.

[0235] Diagnostic information and question-and-answer data are input into the query model to retrieve medical knowledge base information and obtain suggested information.

[0236] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0237] Obtain standard question-and-answer data corresponding to standard medical record reports;

[0238] Standard question-and-answer data is input into the candidate diagnostic model for training to obtain the preset diagnostic model.

[0239] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0240] Obtain first-sample question-and-answer data from multiple first-sample users during their consultation process within a historical time period;

[0241] Each first sample question and answer data is input into the candidate diagnostic model for diagnosis and identification, resulting in multiple candidate medical record reports;

[0242] The medical record reports that meet the preset conditions are selected from multiple candidate medical record reports and used as the standard medical record reports;

[0243] The first sample question and answer data corresponding to the standard medical record report is selected from the first sample question and answer data and used as the standard question and answer data.

[0244] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0245] Acquire second-sample question-and-answer data from multiple second-sample users during their consultations within a historical time period; the second-sample question-and-answer data includes medical question-and-answer data and medical knowledge data.

[0246] The second sample question-and-answer data is input into the initial diagnostic model for training to obtain the candidate diagnostic model.

[0247] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0248] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0249] The technical features of the above embodiments can be combined in any way. For the sake of brevity, 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 application.

[0250] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for generating medical record reports, characterized in that, The method is applied to a diagnostic device, and the method includes: Obtain the standard question-and-answer data corresponding to the standard medical record report, including: Obtain first-sample question-and-answer data from multiple first-sample users during their consultation process within a historical time period; Each of the first sample question-and-answer data is input into the candidate diagnostic model for diagnostic identification, resulting in multiple candidate medical record reports; The medical record reports that meet the preset conditions are selected from the multiple candidate medical record reports and used as the standard medical record reports; The first sample question and answer data corresponding to the standard medical record report is selected from the first sample question and answer data and used as the standard question and answer data. The standard question-and-answer data is input into the candidate diagnostic model for training to obtain a preset diagnostic model; Obtain the user's question and answer data during the current consultation process; The question-and-answer data is input into the preset diagnostic model for diagnostic identification to obtain a diagnostic result; the diagnostic result includes diagnostic information and suggestion information; the preset diagnostic model is trained based on sample question-and-answer data and an initial diagnostic model. Based on the question-and-answer data and the diagnostic results, a medical record report is generated; The preset diagnostic model includes an identification network, a prediction network, and a query network; the step of inputting the question-and-answer data into the preset diagnostic model for diagnostic identification and obtaining diagnostic results includes: The question-and-answer data is input into the identification network for disease type identification to obtain the user's disease type. Then, the disease type and the question-and-answer data are input into the prediction network for prediction to obtain the user's diagnostic information. The diagnostic information is then input into the query network to obtain the corresponding suggestion information. Finally, the diagnostic information and the suggestion information are output. The diagnostic information includes the location and severity of the user's symptoms. The suggestion information refers to treatment recommendations provided to the user. The step of selecting medical record reports that meet preset conditions from the plurality of candidate medical record reports as the standard medical record report includes: Each candidate medical record report is manually scored to obtain a score for each candidate medical record report; the evaluation criteria for manual scoring include the completeness, accuracy, logic, and compliance with medical document standards of the medical record. Candidate medical record reports that meet the preset score threshold will be used as the standard medical record reports.

2. The method according to claim 1, characterized in that, The acquisition of user question-and-answer data during the current consultation includes: Obtain the user's question data during the current consultation process; The question data is input into the response model for identification to obtain the response data during the consultation process; The question-and-answer data is generated based on the question data and the answer data.

3. The method according to claim 1, characterized in that, The method further includes: Acquire second-sample question-and-answer data from multiple second-sample users during their consultations within a historical time period; the second-sample question-and-answer data includes medical question-and-answer data and medical knowledge data. The second sample question-and-answer data is input into the initial diagnostic model for training to obtain the candidate diagnostic model.

4. The method according to claim 1, characterized in that, The step of generating a medical record report based on the question-and-answer data and the diagnostic results includes: Obtain medical record report templates; The question-and-answer data and the diagnosis results are input into the corresponding positions in the medical record report template to obtain the medical record report.

5. A medical record report generation device, characterized in that, Applied to diagnostic equipment, the device includes: The first acquisition module is used to acquire standard question-and-answer data corresponding to standard medical record reports, including: The first acquisition unit is specifically used to acquire first sample question and answer data of multiple first sample users during the consultation process within a historical time period. The first identification unit is specifically used to input the question-and-answer data of each of the first samples into the candidate diagnostic model for diagnosis and identification, and to obtain multiple candidate medical record reports. The report determination unit is specifically used to select medical record reports that meet preset conditions from the plurality of candidate medical record reports as the standard medical record reports; The data determination unit is specifically used to filter out the first sample question and answer data corresponding to the standard medical record report from the first sample question and answer data, and use it as the standard question and answer data. The first training unit is specifically used to input the standard question-and-answer data into the candidate diagnostic model for training, so as to obtain a preset diagnostic model. The second acquisition module is used to acquire the user's question and answer data during the current consultation process; The identification module is used to input the question-and-answer data into the preset diagnostic model for diagnostic identification and to obtain a diagnostic result; the diagnostic result includes diagnostic information and suggestion information; the preset diagnostic model is trained based on sample question-and-answer data and an initial diagnostic model. A generation module is used to generate a medical record report based on the question-and-answer data and the diagnostic results; The preset diagnostic model includes an identification network, a prediction network, and a query network. The identification module is further configured to input the question-and-answer data into the identification network for disease type identification to obtain the user's disease type; then input the disease type and the question-and-answer data into the prediction network for prediction to obtain the user's diagnostic information; finally, input the diagnostic information into the query network to obtain corresponding suggestion information; and output the diagnostic information and the suggestion information. The diagnostic information includes the user's symptom location and severity; the suggestion information refers to treatment recommendations provided to the user. The report determination unit is further configured to manually score each of the candidate medical record reports to obtain a score for each candidate medical record report; the evaluation criteria for manual scoring include the completeness, accuracy, logic, and compliance with medical document standards of the medical record; and the candidate medical record reports that meet the preset score threshold are used as the standard medical record reports.

6. The apparatus according to claim 5, characterized in that, The acquisition module includes: The second acquisition unit is specifically used to acquire the user's question data during the current consultation process; The second identification unit is specifically used to input the question data into the response model for identification, so as to obtain the response data in the consultation process; The generation unit is specifically used to generate the question-and-answer data based on the question data and the answer data.

7. The apparatus according to claim 5, characterized in that, The first acquisition module further includes: The third acquisition unit is specifically used to acquire second sample question-and-answer data of multiple second sample users during the consultation process within a historical time period; the second sample question-and-answer data includes medical question-and-answer data and medical knowledge data. The second training unit is specifically used to input the second sample question-and-answer data into the initial diagnostic model for training, so as to obtain the candidate diagnostic model.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 4.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

Images