Special disease database filling method and device, electronic equipment, storage medium and product

Abstract

This application proposes a method, apparatus, electronic device, storage medium, and product for filling in a disease-specific database. The method, based on a pre-built medical instruction knowledge base and pre-collected medical data of patients with specific diseases, determines data extraction instructions corresponding to the items to be filled in the disease-specific database. The medical instruction base includes data extraction instruction templates corresponding to the items to be filled. By executing the data extraction instructions, data corresponding to the items to be filled in the disease-specific database is extracted from the medical data of patients with specific diseases. The data is then filled into the items to be filled in the disease-specific database. Using the technical solution of this application, the automatic acquisition and automatic filling of data corresponding to the items to be filled in the disease-specific database can be achieved, improving the efficiency and accuracy of disease-specific database filling.

Description

Technical Field

[0001] This application relates to the field of data extraction technology, and in particular to a method, apparatus, electronic device, storage medium and product for filling in a disease database. Background Technology

[0002] In the healthcare industry, the establishment and maintenance of disease-specific databases are crucial for disease surveillance, epidemiological research, and public health management. Current methods for filling out disease-specific databases rely on manual input; that is, medical personnel manually enter patients' clinical information, diagnoses, treatment processes, and follow-up data. While manual data entry ensures the specificity and detail of the data, it is inefficient and prone to errors, especially with large datasets. The workload and error rate of manual processing become the main factors limiting efficiency, resulting in low efficiency and accuracy in disease-specific database entry. Summary of the Invention

[0003] Based on the above needs, this application proposes a method, device, electronic device, storage medium, and product for filling in a disease database, which can improve the efficiency and accuracy of filling in the disease database.

[0004] To achieve the above objectives, this application proposes the following technical solution:

[0005] According to a first aspect of the embodiments of this application, a method for filling in a disease database is provided, including:

[0006] Based on a pre-built medical instruction library and pre-collected medical data of patients with specific diseases, data extraction instructions corresponding to the items to be populated in the specific disease library are determined; the medical instruction library includes data extraction instruction templates corresponding to the items to be populated.

[0007] By executing the data extraction instruction, data corresponding to the items to be filled in the disease database is extracted from the medical data of the disease patients.

[0008] The data is then populated into the fields to be populated in the disease database.

[0009] Optionally, based on a pre-built medical instruction library and pre-collected medical data of patients with specific diseases, data extraction instructions corresponding to the items to be populated in the disease-specific library are determined, including:

[0010] According to the preset instruction mapping rules, the data extraction instruction template corresponding to the items to be filled in the disease database is determined from the pre-built medical instruction library; the instruction mapping rules include the correspondence between the items to be filled and the data extraction instruction template.

[0011] According to the preset medical record fragment mapping rules, the source of medical record fragments corresponding to the items to be filled in the disease database is determined from the pre-collected medical data of patients with specific diseases; the medical record fragment mapping rules include the correspondence between the items to be filled and the sources of medical record fragment extraction.

[0012] Based on the medical record fragment extraction source corresponding to the item to be populated in the disease database and the data extraction instruction template corresponding to the item to be populated in the disease database, a data extraction instruction corresponding to the item to be populated in the disease database is constructed.

[0013] Optionally, based on the medical record fragment extraction source corresponding to the item to be populated in the disease database and the data extraction instruction template corresponding to the item to be populated in the disease database, a data extraction instruction corresponding to the item to be populated in the disease database is constructed, including:

[0014] The source of the extracted medical record fragments corresponding to the items to be filled in the disease database is filled into the data extraction instruction template corresponding to the items to be filled in the disease database, so as to obtain the data extraction instruction corresponding to the items to be filled in the disease database.

[0015] Optionally, by executing the data extraction instruction, data corresponding to the items to be populated in the disease database is extracted from the disease-specific patient medical data, including:

[0016] According to the preset medical record fragment mapping rules, extract the medical record fragments corresponding to the items to be filled in the disease database from the medical data of the disease patients;

[0017] By executing the data extraction instruction, the data corresponding to the items to be filled in the disease database is extracted from the medical record fragments corresponding to the items to be filled in the disease database.

[0018] Optionally, if the item to be populated in the disease database has a dependency relationship with other items to be populated in the disease database, then the item to be populated in the disease database and other items to be populated in the disease database that have a dependency relationship with the item to be populated in the disease database correspond to the same data extraction instruction;

[0019] By executing the data extraction instruction, data corresponding to the items to be populated in the disease database is extracted from the medical data of the disease patients, including:

[0020] By executing the data extraction instruction, data corresponding to the items to be filled in the disease database and data corresponding to other items to be filled in the disease database that have a dependency relationship with the items to be filled in the disease database are extracted from the medical data of the disease patients.

[0021] Optionally, by executing the data extraction instruction, data corresponding to the items to be populated in the disease database is extracted from the disease-specific patient medical data, including:

[0022] The data extraction instructions for the items to be filled in the disease database and the medical data of the patients with the disease are input into the pre-built data extraction model to obtain the data to be filled in the items;

[0023] The data extraction model is used to extract data corresponding to the items to be filled in the disease database from the medical data of the disease patients by executing the data extraction instructions; wherein, the data extraction model is a large language model.

[0024] Optionally, before populating the data into the disease database to be populated items, the method further includes:

[0025] According to the pre-set quality inspection rules, the data corresponding to the items to be filled in the disease database is inspected to determine the quality inspection information corresponding to the data.

[0026] If the quality inspection information indicates that the quality inspection is qualified, then the data will be filled into the items to be filled in the disease database.

[0027] If the quality inspection information indicates that the quality inspection is unqualified, a prompt message indicating that the data output is incorrect will be output.

[0028] Optionally, the methods for filling in the disease database also include:

[0029] If the quality inspection information indicates that it is unqualified, then the training sample corresponding to the item to be filled in the disease database and the real data corresponding to the training sample are collected; wherein, the training sample corresponding to the item to be filled in the disease database includes: the data extraction instruction corresponding to the item to be filled in the disease database and the sample medical data corresponding to the item to be filled in the disease database.

[0030] The training samples are input into the data extraction model to obtain the predicted data corresponding to the training samples;

[0031] Based on the difference between the real data corresponding to the training samples and the predicted data corresponding to the training samples, the parameters of the data extraction model are adjusted.

[0032] Optionally, before determining the data extraction instructions corresponding to the items to be populated in the disease database based on a pre-built medical instruction library and pre-collected medical data of patients with specific diseases, the process may also include:

[0033] Identify the forms to be filled in the disease database and extract at least one item to be filled in the disease database from the forms.

[0034] According to a second aspect of the embodiments of this application, a disease database filling device is provided, comprising:

[0035] The instruction determination module is used to determine the data extraction instruction corresponding to the item to be populated in the disease database based on a pre-built medical instruction library and pre-collected medical data of patients with specific diseases; the medical instruction library includes data extraction instruction templates corresponding to the items to be populated.

[0036] The data determination module is used to extract the data corresponding to the items to be filled in the disease database from the medical data of the disease patients by executing the data extraction instruction;

[0037] The data entry module is used to fill the data into the items to be filled in the disease database.

[0038] According to a third aspect of the embodiments of this application, an electronic device is provided, including: a memory and a processor;

[0039] The memory is connected to the processor and is used to store programs;

[0040] The processor is used to implement the above-mentioned disease database filling method by running the program in the memory.

[0041] According to a fourth aspect of the embodiments of this application, a storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the above-described method for filling in the disease database is implemented.

[0042] According to a fifth aspect of the embodiments of this application, a computer program product is provided, including computer program instructions, which, when executed by a processor, cause the processor to implement the above-described disease database filling method.

[0043] The proposed method for filling in a disease-specific database in this application is based on a pre-built medical instruction knowledge base and pre-collected medical data of patients with specific diseases. It determines the data extraction instructions corresponding to the items to be filled in the disease-specific database. The medical instruction base includes data extraction instruction templates corresponding to the items to be filled. By executing the data extraction instructions, the data corresponding to the items to be filled in the disease-specific database is extracted from the medical data of patients with specific diseases. The data is then filled into the items to be filled in the disease-specific database. Using the technical solution of this application, the automatic acquisition and automatic filling of data corresponding to the items to be filled in the disease-specific database can be achieved, improving the efficiency and accuracy of disease-specific database filling. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0045] Figure 1 A flowchart illustrating a method for filling in a disease database, provided as an embodiment of this application;

[0046] Figure 2 This is a schematic diagram of the processing flow for extracting data corresponding to items to be populated in the disease database, provided in an embodiment of this application.

[0047] Figure 3 This is a schematic diagram of the data quality inspection process for the items to be populated in the disease database, provided in an embodiment of this application.

[0048] Figure 4 This is a schematic diagram of the data extraction model training process provided in an embodiment of this application;

[0049] Figure 5 A schematic diagram of the structure of a disease database filling device provided in this application embodiment;

[0050] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0051] The technical solution of this application is applicable to data extraction scenarios, specifically for data extraction in filling out disease-specific databases. By adopting the technical solution of this application, the automatic acquisition and automatic filling of data for items to be filled in the disease-specific database can be achieved, improving the efficiency and accuracy of disease-specific database filling.

[0052] A disease-specific database, also known as a disease-specific specialty database or single-disease database, is a standardized and regulated database containing clinical data of patients with a specific disease or condition. The purpose of establishing such a database is to collect and integrate comprehensive information on a particular disease or condition, including multi-dimensional data such as patient biometrics, genotype, lifestyle, medication use, and laboratory test results, in order to facilitate in-depth disease research and clinical studies.

[0053] In the healthcare industry, the establishment and maintenance of disease-specific databases are crucial for disease surveillance, epidemiological research, and public health management. Traditional methods of filling out disease-specific databases rely on manual input, where medical personnel manually enter patients' clinical information, diagnoses, treatment processes, and follow-up data. While manual data entry ensures the specificity and detail of the data, it is inefficient, especially with large datasets, being time-consuming and labor-intensive. Furthermore, manual entry is prone to errors, such as input errors, omissions of important information, or data inconsistencies. These problems can lead to decreased data quality and lower accuracy in disease-specific database entries, impacting the accuracy of subsequent data analysis and research results.

[0054] Based on this, this application proposes a method for filling in a disease database. This technical solution can automatically acquire and fill in the data for items to be filled in the disease database, thereby solving the problem of low efficiency and accuracy of disease database filling in the prior art.

[0055] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0056] Exemplary methods

[0057] See Figure 1 As shown in the embodiment of this application, a method for filling in a disease database is proposed. The method includes:

[0058] S101. Based on the pre-built medical instruction library and the pre-collected medical data of patients with specific diseases, determine the data extraction instructions corresponding to the items to be populated in the specific disease library.

[0059] Specifically, for each disease-specific database, there is a corresponding disease-specific form that needs to be filled out. This form serves as the "to-fill" form for the disease-specific database. The to-fill form for each disease-specific database lists the items that need to be filled in when submitting the data for that disease. For example, the to-fill form might include clinical information, diagnosis, treatment process, and follow-up data for the patient. For example, the to-fill form for the coronary heart disease-specific database might include: age, gender, and occupation for clinical information; family history, past medical history, and comorbidities for diagnosis; blood pressure, heart rate, weight, height, blood lipids, blood glucose, anesthesia method, and medication information for treatment; and follow-up data for follow-up, including readmissions, dietary habits, and exercise habits. Therefore, this embodiment first needs to extract the to-fill items for each disease-specific database from its to-fill form.

[0060] In this embodiment, after extracting the items to be populated from the disease-specific database corresponding to the symptoms, it is necessary to determine the data extraction instructions corresponding to the items to be populated based on a pre-constructed medical instruction library and pre-collected medical data of patients with specific diseases. The medical instruction library includes data extraction instruction templates corresponding to the items to be populated. In this embodiment, each data extraction instruction template corresponding to an item in the medical instruction library is designed by medical personnel with professional medical knowledge. Therefore, the data extraction instructions determined using these templates are highly professional and scientific. The data extraction instruction templates corresponding to the items to be populated in this embodiment are designed based on the data extraction intent and data output standards of the items to be populated; that is, the template includes the data extraction intent and data output standards for the items to be populated. Furthermore, the medical instruction library constructed in this embodiment can include data extraction instruction templates corresponding to items to be populated from several disease-specific databases, or a separate medical instruction library can be constructed for each disease-specific database, with each database containing only the data extraction instruction templates corresponding to the items to be populated in that specific database.

[0061] In this embodiment, the pre-collected medical data for patients with specific diseases are medical record data collected from the hospital's electronic medical record system or physical examination data collected from the system of a physical examination hospital. That is, the medical data includes medical record data and physical examination data. For example, the medical record data included in the hospital's electronic medical record system covers the patient's basic information, admission and discharge records, detailed medical history, diagnostic information, treatment information, laboratory test results, imaging examination results, medication prescriptions, and clinician notes. Basic information typically includes the patient's name, ID number, date of birth, occupation, etc., and is usually recorded in the admission and / or discharge records. The admission record also includes the admission time, reason for admission, and preliminary diagnosis; the discharge record also includes the discharge date, treatment results, and discharge instructions. The medical history records the patient's past medical history, family medical history, and chronic disease status in detail. Laboratory test results cover various blood tests, urine tests, and other laboratory results, while imaging examination results include X-rays, CT scans, MRI, etc. Medication prescriptions record the patient's medication types, dosages, and administration times in detail. Clinicians' notes include treatment plans, surgical records, and follow-up information, providing patients with comprehensive medical records and ongoing health management.

[0062] This embodiment describes the entry of a disease-specific database for a particular condition. In the pre-collected medical data of patients with the disease, the patients are those diagnosed with the disease or those with certain symptoms of the disease. For example, when entering the disease-specific database for coronary heart disease, the pre-collected medical data of patients with the disease includes medical data of patients diagnosed with coronary heart disease or medical data of patients with symptoms of coronary heart disease (e.g., medical data of patients with chest pain symptoms).

[0063] In this embodiment, the medical data collected for patients with specific diseases not only includes the medical data of the patients with specific diseases, but also records the position of each medical data point within the medical data of all patients with specific diseases. This position serves as the extraction source when extracting medical data, i.e., the source for extracting medical record fragments. For example, if the item to be filled in the specific disease database is age, and the position of age within the medical data of all patients with specific diseases might be the age recorded in the admission record, then the source for extracting the medical record fragment for age would be the admission record - age.

[0064] In this embodiment, a data extraction instruction template corresponding to the item to be populated in the disease database can be obtained from a pre-built medical instruction library, and the source of the medical record fragment corresponding to the item to be populated in the disease database can be obtained from pre-collected medical data of patients with specific diseases. Based on the obtained data extraction instruction template and the source of the medical record fragment, a data extraction instruction corresponding to the item to be populated can be constructed. This data extraction instruction contains the intent for extracting data corresponding to the item to be populated and the data output standard. The specific steps are as follows:

[0065] First, according to the preset instruction mapping rules, the data extraction instruction template corresponding to the items to be populated in the disease database is determined from the pre-built medical instruction library.

[0066] This embodiment pre-sets instruction mapping rules, which include the correspondence between the items to be populated in the disease-specific database and the data extraction instruction templates. According to these rules, the data extraction instruction template corresponding to the item to be populated in the disease-specific database can be retrieved from all the data extraction instruction templates stored in the pre-built medical instruction database. Since the items to be populated in the disease-specific databases differ for different diseases, the correspondence in the instruction mapping rules for different disease-specific databases also differs. Therefore, different instruction mapping rules need to be set for different disease-specific databases when setting up the instruction mapping rules.

[0067] Second, according to the preset medical record fragment mapping rules, the source of the medical record fragments corresponding to the items to be filled in the disease database is determined from the pre-collected medical data of patients with specific diseases.

[0068] This embodiment pre-sets medical record fragment mapping rules, which include the correspondence between the items to be filled in the disease-specific database and the sources from which the medical record fragments are extracted. According to these rules, the sources from which the medical record fragments corresponding to the items to be filled in the disease-specific database can be determined from the pre-collected medical data of patients with specific diseases. Since the items to be filled in the disease-specific databases differ for different diseases, the correspondence in the medical record fragment mapping rules for different disease-specific databases also differs. Therefore, when setting the medical record fragment mapping rules, different rules need to be set for different disease-specific databases. For example, Table 1 below records part of the content of the medical record fragment mapping rules corresponding to the coronary heart disease disease database:

[0069] Table 1

[0070]

[0071]

[0072] Third, based on the medical record fragment extraction source corresponding to the items to be populated in the disease database and the data extraction instruction template corresponding to the items to be populated in the disease database, data extraction instructions for the items to be populated in the disease database are constructed.

[0073] In this embodiment, after determining the source of medical record fragment extraction and the data extraction instruction template corresponding to the item to be populated in the disease database, the data extraction instruction corresponding to the item to be populated in the disease database is constructed by combining the source of medical record fragment extraction and the data extraction instruction template. Specifically, the combination of the source of medical record fragment extraction and the data extraction instruction template can be achieved by recombining all the contents of the data extraction instruction template and all the contents of the source of medical record fragment extraction in the disease database, resulting in a data extraction instruction that includes both the contents of the data extraction instruction template and the contents of the source of medical record fragment extraction.

[0074] Furthermore, the combination of the medical record fragment extraction source and the data extraction instruction template corresponding to the item to be filled in the disease database can also be as follows: The medical record fragment extraction source corresponding to the item to be filled in the disease database is filled into the data extraction instruction template corresponding to the item to be filled in the disease database, resulting in the data extraction instruction corresponding to the item to be filled in the disease database. That is, the data extraction instruction template for the item to be filled in the disease database pre-sets the filling position for the medical record fragment extraction source. The medical record fragment extraction source corresponding to the item to be filled in the disease database is directly filled into the filling position set in the data extraction instruction template, thus obtaining a complete data extraction instruction. The data extraction instruction corresponding to the item to be filled can play a guiding role when extracting data corresponding to the item to be filled, extracting data that conforms to the data extraction intent in the data extraction instruction and outputting the data corresponding to the item to be filled according to the data output standard in the data extraction instruction, so that the final data output format of the data corresponding to the item to be filled conforms to the data output standard in the data extraction instruction. For example, when the item to be filled is age, the data extraction instruction corresponding to this item is:

[0075] "Now you are a hospital data analyst, and you need to start from a..." Admission Record - Age Extract the patients' ages for statistical analysis. Specific requirements are as follows:

[0076] 1. Focus on accurately extracting age information from the text, which usually follows the "Age:" label;

[0077] 2. Age should be extracted as a complete string with units (e.g., "42 years old") to ensure unit consistency and facilitate subsequent processing;

[0078] 3. Handle potential anomalies, such as missing age fields or format errors (e.g., non-numeric characters). Extracted numerical values ​​should be between 0 and 150.

[0079] In the example of the data extraction instruction corresponding to the above-mentioned item to be filled, the position of "Admission Record - Age" is the fill position set in the data extraction instruction template. "Admission Record - Age" is the source of the medical record fragment extraction corresponding to the item to be filled. All other content except "Admission Record - Age" is the data extraction instruction template corresponding to the item to be filled.

[0080] S102. By executing the data extraction command, extract the data corresponding to the items to be filled in the disease database from the medical data of patients with specific diseases.

[0081] After determining the data extraction instructions corresponding to the items to be filled in the disease database through the above steps, the data extraction instructions are executed. Following the data extraction intent contained in the instructions and the source of the medical record fragments corresponding to the items to be filled in the disease database, the data corresponding to the items to be filled in the disease database is extracted from the patient's medical data at the position corresponding to the source of the medical record fragments. This data is then output according to the data output standards in the data extraction instructions, ensuring that the data output format of the data corresponding to the items to be filled conforms to the data output standards in the data extraction instructions. In this embodiment, the data extraction instructions for the items to be filled in the disease database are constructed based on the data extraction instruction template and the source of the medical record fragments corresponding to the items to be filled in the disease database. Because the data extraction instruction template is designed by medical personnel with professional medical knowledge and possesses a high degree of professionalism and scientific rigor, the data quality of the data corresponding to the items to be filled in the disease database extracted using the data extraction instructions is higher, i.e., the data accuracy is higher. For example, when the item to be filled in the disease database is age, the output data corresponding to the item to be filled in the disease database can be as follows:

[0082] {

[0083] “patient_age”:“42”,

[0084] "patient_age_unit": "years old"

[0085] }

[0086] Specifically, in this embodiment, a data extraction model is pre-built. The data extraction instructions corresponding to the items to be filled in the disease database and the pre-collected medical data of disease patients are input into this pre-built data extraction model, enabling the model to execute the data extraction instructions and extract the data corresponding to the items to be filled in the disease database from the medical data of disease patients. Preferably, this data extraction model is a Large Language Model (LLM). The deep learning capabilities of LLM can accurately parse information from complex medical data, and its high-level language understanding capabilities can handle data of various formats and sources, exhibiting high adaptability. This allows the disease database filling method of this embodiment to be effectively deployed in different medical environments without requiring system redesign or adjustment for each specific situation, greatly improving the application flexibility of the disease database filling method.

[0087] S103. Populate the data into the items to be populated in the disease database.

[0088] In this embodiment, after extracting the data corresponding to the items to be filled in the disease database through the above steps, it is necessary to fill the corresponding data of the items to be filled in the disease database into the corresponding fill positions in the disease database. That is, the data of the items to be filled in the disease database is filled into the items to be filled in the form of the disease database.

[0089] In this embodiment, data for patients with specific diseases needs to be populated into the disease database individually; that is, data is populated for each patient one by one. Specifically, medical data for patients with specific diseases is collected in advance, i.e., the medical data for one patient with a specific disease is collected in advance. Then, by executing a data extraction command, the data corresponding to the items to be populated in the disease database is extracted from the medical data of one patient with a specific disease and populated into the items to be populated in the disease database, thus completing the disease database population for one patient. Then, the medical data of another patient with a specific disease is collected to complete the disease database population for that patient, until finally, the data corresponding to the items to be populated in the disease database for all patients with specific diseases are populated into the items to be populated in the disease database.

[0090] In this embodiment, a disease database contains at least one disease database item to be populated. When a disease database contains multiple disease database items to be populated, data extraction instructions corresponding to multiple disease database items to be populated can be determined simultaneously, thereby extracting the data corresponding to multiple disease database items to be populated at the same time, and populating the data corresponding to multiple disease database items to be populated into the disease database items to be populated at the same time.

[0091] As described above, the disease-specific database filling method proposed in this application, based on a pre-built medical instruction knowledge base and pre-collected medical data of disease-specific patients, determines the data extraction instruction corresponding to the item to be filled in the disease-specific database. The medical instruction base includes data extraction instruction templates corresponding to the items to be filled. By executing the data extraction instruction, the data corresponding to the item to be filled in the disease-specific database is extracted from the medical data of disease-specific patients. The data is then filled into the item to be filled in the disease-specific database. Using the technical solution of this embodiment, the automatic acquisition of data corresponding to the item to be filled in the disease-specific database and the automatic filling of data corresponding to the item to be filled in the disease-specific database can be achieved, improving the efficiency and accuracy of disease-specific database filling.

[0092] As an optional implementation, see [link to implementation details]. Figure 2 As shown, in another embodiment of this application, step S102, which involves extracting data corresponding to the items to be filled in the disease database from the medical data of disease patients by executing a data extraction instruction, specifically includes the following steps:

[0093] S201. According to the preset medical record fragment mapping rules, extract the medical record fragments corresponding to the items to be filled in the disease database from the medical data of patients with specific diseases.

[0094] In this embodiment, before extracting the data corresponding to the items to be filled in the disease database, the medical record fragments corresponding to the items to be filled in the disease database can be extracted from the pre-collected medical data of patients with specific diseases according to the preset medical record fragment mapping rules. Specifically, the source of the medical record fragment extraction corresponding to the items to be filled in the disease database is first determined according to the preset medical record fragment mapping rules. Then, according to the source of the medical record fragment extraction corresponding to the items to be filled in the disease database, the medical record fragments corresponding to the source of the medical record fragment extraction are extracted from the pre-collected medical data of patients with specific diseases as the medical record fragments corresponding to the items to be filled in the disease database.

[0095] For example, when the item to be populated in the disease database is age, the source of the medical record fragment corresponding to this item is the admission record. Therefore, the medical record fragment corresponding to the item to be populated in the disease database extracted from the medical data of patients with specific diseases is the medical record fragment of the admission record, which is shown below:

[0096] Patient's name: Zhang San

[0097] Age: 42

[0098] Gender: Male

[0099] Admission date: April 12, 2024

[0100] Chief complaint: Chest pain lasting 3 hours

[0101] Diagnosis: Angina pectoris

[0102] S202. By executing the data extraction command, extract the data corresponding to the items to be filled in the disease database from the medical record fragments corresponding to the items to be filled in the disease database.

[0103] This embodiment extracts the corresponding data for the disease-specific database entries directly from the medical record fragments extracted from the disease-specific patient medical data by executing data extraction commands. Compared to directly extracting the corresponding data from the disease-specific patient medical data by executing data extraction commands, first extracting the corresponding medical record fragments from the disease-specific patient medical data and then extracting the corresponding data from those fragments results in higher accuracy in the extracted data.

[0104] Specifically, when using the data extraction model to extract data corresponding to the items to be filled in the disease database, the data extraction command corresponding to the items to be filled in the disease database and the medical record fragments corresponding to the items to be filled in the disease database extracted from the medical data of patients with specific diseases are directly input into the data extraction model. The data extraction model extracts the data corresponding to the items to be filled in the disease database from the medical record fragments corresponding to the items to be filled in the disease database by executing the data extraction command. Compared with using the data extraction model to execute the data extraction command to extract the data corresponding to the items to be filled in the disease database from the medical data of patients with specific diseases, the data extraction model in this embodiment has higher processing efficiency and the extracted data corresponding to the items to be filled in the disease database is more accurate.

[0105] As an optional implementation, another embodiment of this application discloses that among the disease database entries to be filled in the disease database form, some may have dependencies. If a disease database entry has no other dependent entries, then that entry corresponds to a separate data extraction instruction. If a disease database entry has dependencies on other disease database entries, then that entry and the other dependent entries correspond to the same data extraction instruction. In this case, step S102, extracting the data corresponding to the disease database entry from the disease patient's medical data by executing the data extraction instruction, includes:

[0106] By executing data extraction commands, data corresponding to items to be populated in the disease database and data corresponding to other disease database items that are dependent on the items to be populated in the disease database are extracted from the medical data of patients with specific diseases.

[0107] Specifically, by using the same data extraction command corresponding to the item to be populated in the disease-specific database and other items in the disease-specific database that have dependencies on it, data corresponding to the item to be populated in the disease-specific database and other items in the disease-specific database that have dependencies on it can be extracted from the medical data of patients with the specific disease. For example, item A in the disease-specific database has dependencies on both item B and item C in the disease-specific database, meaning that other items in the disease-specific database that have dependencies on item A are item B and item C. In this case, item A, item B, and item C in the disease-specific database correspond to the same data extraction command. By executing this single data extraction command, data corresponding to item A, item B, and item C in the disease-specific database can be extracted from the medical data of patients with the specific disease.

[0108] If a disease-specific database item to be populated has dependencies on other disease-specific database items to be populated, then the pre-built medical instruction library contains the same data extraction instruction template corresponding to the item to be populated and the other disease-specific database items that depend on it. The method for determining the same data extraction instruction template corresponding to the item to be populated and the other disease-specific database items that depend on it is as follows: First, according to preset instruction mapping rules, determine the same data extraction instruction template corresponding to the item to be populated and the other disease-specific database items that depend on it from the pre-built medical instruction library; then, according to preset medical record fragment mapping rules, determine the medical record fragment extraction source corresponding to the item to be populated and the medical record fragment extraction source corresponding to the other disease-specific database items that depend on it from the pre-collected medical data of disease-specific patients; finally, based on the same data extraction instruction template corresponding to the item to be populated and the other disease-specific database items that depend on it,... The template includes the source of medical record fragment extraction for the item to be filled in the disease database and the source of medical record fragment extraction for other disease database items that are dependent on the item to be filled in the disease database. It constructs the same data extraction instruction for the item to be filled in the disease database and the other disease database items that are dependent on the item to be filled in the disease database. Specifically, it fills the source of medical record fragment extraction for the item to be filled in the disease database and the other disease database items that are dependent on the item to be filled in the disease database into the same data extraction instruction template for the item to be filled in the disease database and the other disease database items that are dependent on the item to be filled in the disease database, thus obtaining the same data extraction instruction for the item to be filled in the disease database and the other disease database items that are dependent on the item to be filled in the disease database.

[0109] For example, in actual medical practice, it is necessary to first determine whether there are any narrowed coronary arteries. If so, it is then necessary to determine which specific coronary arteries are narrowed (left main coronary artery, right coronary artery, left anterior descending artery, etc.), the degree of narrowing, the presence of thrombosis, whether balloon angioplasty was used, whether a stent was placed, and whether the stent placement was successful. Therefore, it can be seen that the patient position, puncture site, name of the narrowed vessel, degree of stenosis, whether balloon angioplasty was performed, whether a stent was placed, and whether the stent placement was successful are all dependent on each other in the disease database. In this case, when the disease database entries are patient position, puncture site, name of the narrowed vessel, degree of stenosis, whether balloon angioplasty was performed, whether a stent was placed, and whether the stent placement was successful, the data extraction instruction is as follows:

[0110] "You are a medical records administrator working in a large hospital. Your main responsibility is to organize and record surgical procedures and ensure that all documents are accurate for subsequent medical staff to refer to."

[0111] Specific task requirements:

[0112] 1. Focus on the input Surgical Record - Surgical Procedure If other information is entered along with the data, you don't need to pay attention to it.

[0113] 2. Output the content of the fields of interest (patient position, puncture site, name of the stenotic vessel, degree of stenosis, whether balloon dilation was performed, whether a stent was placed, and whether the stent placement was successful) in JSON format;

[0114] 3. These fields have fixed dependencies: the "degree of stenosis" can only be described if the "name of the stenotic vessel" exists; the "whether a stent was placed" must be "yes" to describe whether the stent placement was successful. If the prerequisite is missing, the field content still needs to be recorded, such as "no stenotic vessel exists, so the degree of stenosis cannot be described", so that the output JSON is completely fixed and easy to parse and process later.

[0115] 4. If Surgical Record - Surgical Procedure If the content of the fields of interest is not described, simply record "Not described";

[0116] 5. Output the JSON content directly; do not provide any further description.

[0117] The above-mentioned "Surgical Record - Surgical Process" is the source of medical record fragments for extraction, including patient position, puncture site, name of stenotic vessel, degree of stenosis, whether balloon dilation was performed, whether a stent was placed, and whether the stent placement was successful. The other contents in the above data extraction instructions, excluding "Surgical Record - Surgical Process", are the data extraction instruction templates for patient position, puncture site, name of stenotic vessel, degree of stenosis, whether balloon dilation was performed, whether a stent was placed, and whether the stent placement was successful.

[0118] Correspondingly, by executing the above data extraction command, the data corresponding to the items to be populated in the disease database extracted from the medical data of disease patients is as follows:

[0119] {

[0120] "Patient Position": "Supine position"

[0121] Puncture site: Right radial artery

[0122] “Disease-causing blood vessels”: [{

[0123] “Name of the stenotic vessel”: “Left Anterior Descending Artery (LAD)”

[0124] "Narrowness": 75%

[0125] }]

[0126] "Balloon dilation required?": "Yes"

[0127] "Stent placement required": "Yes"

[0128] "Was stent placement successful?": "Yes"

[0129] }

[0130] Correspondingly, when the data corresponding to the disease-specific database items to be populated is extracted from the disease-specific patient medical data by executing the above data extraction instructions, and the medical record fragments corresponding to the disease-specific database items to be populated are extracted from the disease-specific patient medical data according to the preset medical record fragment mapping rules, the extracted medical record fragments corresponding to the disease-specific database items to be populated, that is, the medical record fragments whose source is "surgical record - surgical procedure", are as follows:

[0131] "After local anesthesia, the patient was placed in a supine position and kept stable. A catheter was inserted through a puncture of the right radial artery to the ostium of the coronary artery. Contrast agent was injected, and dynamic X-ray observation confirmed a 75% stenosis in the left anterior descending artery (LAD). A balloon catheter was then inserted to the stenosis site for dilation. The initial dilation was not ideal; the stenosis was not completely opened, and the degree of stenosis was reduced to 60%. Given the poor effect of balloon dilation, it was decided to place a metal stent. An appropriately sized stent was selected, inserted, and accurately placed at the stenosis site, and the stent was deployed to ensure good adhesion to the vessel wall. After stent placement, contrast agent was injected again to confirm that the stent was correctly positioned, the LAD was well opened, the stenosis was relieved, and blood flow returned to normal. All catheter materials were retrieved, and pressure was applied to the puncture site for hemostasis and bandaging. The patient's vital signs were stable during the procedure, and no significant complications occurred. The patient was returned to the ward for continued observation and treatment."

[0132] This embodiment can use a single data extraction command to extract the data corresponding to all dependent items in one go, avoiding multiple calls to the data extraction command. This can significantly improve work efficiency, reduce error rate, and ensure the continuity and accuracy of data processing.

[0133] As an optional implementation, see [link to implementation details]. Figure 3 As shown, in another embodiment of this application, before performing step S103 and filling the data into the disease database to be filled, the following steps are also included:

[0134] S301. According to the pre-set quality inspection rules, perform quality inspection on the data corresponding to the items to be filled in the disease database, and determine the quality inspection information corresponding to the data.

[0135] In this embodiment, quality inspection rules are pre-set for the data corresponding to the items to be populated in the disease database, and different disease database items have different quality inspection rules. For example, when the item to be populated in the disease database is age, the quality inspection rules for the data corresponding to the item to be populated include that the data must include numeric characters, etc.

[0136] This embodiment performs quality inspection on the data corresponding to the extracted disease database items to be filled, according to pre-set quality inspection rules. It determines whether the data for each item conforms to the corresponding quality inspection rules and assigns quality inspection information based on the result. Specifically, if the data for each disease database item conforms to the quality inspection rules, the quality inspection information indicates that the quality inspection is qualified; otherwise, it indicates that the quality inspection is unqualified.

[0137] S302. If the quality inspection information indicates that the quality inspection is qualified, then the data will be filled into the pending items in the disease database.

[0138] If the quality inspection information corresponding to the data indicates that the quality inspection is qualified, the data is populated into the pending items of the disease database, thereby improving the accuracy of the disease database entries. The specific execution method for populating the data into the pending items of the disease database has been described in detail in step S103 of the above embodiment, and will not be repeated in this embodiment.

[0139] S303. If the quality inspection information indicates that the quality inspection is unqualified, output a prompt message indicating that the data output is incorrect.

[0140] If the quality inspection information corresponding to the data indicates that the data is unqualified, an error message will be output to inform the user that the data corresponding to the item to be filled in the disease database is incorrect, and the data for that item should be extracted again. If the data for the item to be filled in the disease database is extracted using a data extraction model, an error message will be output to inform the user that the data extraction model has low accuracy in extracting data for that item, thus allowing for evaluation of the data extraction model. The user can then optimize and train the data extraction model for that specific item in the disease database.

[0141] As an optional implementation, see [link to implementation details]. Figure 4As shown in another embodiment of this application, if the quality inspection information of the data corresponding to the item to be filled in the disease database indicates that the quality inspection is unqualified, or if the data corresponding to the item to be filled in the disease database is manually reviewed and the review result indicates that the data is unqualified, then it is necessary to optimize and train the data extraction model for the item to be filled in the disease database. The specific training steps include:

[0142] S401. Collect the training samples and real data corresponding to the items to be filled in the disease database.

[0143] In this embodiment, when optimizing and training the data extraction model for the items to be filled in the disease database, it is first necessary to collect the training samples corresponding to the items to be filled in the disease database, as well as the real data corresponding to each training sample. The training samples corresponding to the items to be filled in the disease database include: the data extraction instructions corresponding to the items to be filled in the disease database and the sample medical data corresponding to the items to be filled in the disease database. Among them, the data extraction instructions corresponding to the items to be filled in the disease database are determined based on a pre-built medical instruction library and pre-collected medical data of disease patients. The specific steps have been described in detail in step S101 of the above embodiment, and will not be repeated in this embodiment. The sample medical data corresponding to the items to be filled in the disease database can be pre-collected medical data of disease patients, or it can be medical record fragments corresponding to the items to be filled in the disease database extracted from the pre-collected medical data of disease patients according to preset medical record fragment mapping rules. The specific extraction steps have been described in detail in the above embodiment, and will not be repeated in this embodiment. The real data corresponding to the training samples are the real data corresponding to the items to be filled in the disease database in the pre-collected medical data of disease patients.

[0144] S402. Input the training samples into the data extraction model to obtain the prediction data corresponding to the training samples.

[0145] In this embodiment, the collected training samples are input into the data extraction model. This model extracts the data corresponding to the items to be filled in the disease database from the sample medical data of these items in the training samples, using the extracted data as the predicted data for the training samples. The specific execution method of the data extraction model extracting the data corresponding to the items to be filled in the disease database from the sample medical data of these items in the training samples is the same as the execution method of step S102 in the above embodiment, and will not be repeated here.

[0146] S403. Adjust the parameters of the data extraction model based on the difference between the real data corresponding to the training samples and the predicted data corresponding to the training samples.

[0147] This embodiment is based on the difference between the real data corresponding to the training sample and the predicted data corresponding to the training sample. The data extraction model is adjusted with the goal of minimizing the difference between the real data corresponding to the training sample and the predicted data corresponding to the training sample. That is, the loss function between the real data corresponding to the training sample and the predicted data corresponding to the training sample can be calculated. The data extraction model is adjusted with the goal of minimizing the loss function. The optimization training of the data extraction model for the items to be filled in the disease database is stopped when the loss function between the real data corresponding to the training sample and the predicted data corresponding to the training sample reaches the preset standard range.

[0148] As an optional implementation, another embodiment of this application discloses that, before step S101, which determines the data extraction instruction corresponding to the item to be populated in the disease database based on the pre-built medical instruction library and the pre-collected medical data of patients with specific diseases, the following steps are also included:

[0149] Identify the forms to be filled in the disease database and extract at least one item to be filled in the disease database from the forms.

[0150] In this embodiment, each disease-specific database corresponds to a specific disease form that needs to be filled out, and this specific disease form serves as the form to be filled in of the disease database. By recognizing the form to be filled in the disease database, the fields of the items to be filled in the form can be identified, thereby extracting at least one item to be filled in the disease database from the form. The recognition of the form to be filled in the disease database can be performed using existing text recognition methods, such as OCR technology.

[0151] Exemplary device

[0152] Accordingly, this application also provides a disease database filling device, see [link to relevant documentation]. Figure 5 As shown, the device includes:

[0153] The instruction determination module 100 is used to determine the data extraction instruction corresponding to the item to be populated in the disease database based on the pre-built medical instruction library and the pre-collected medical data of patients with specific diseases; the medical instruction library includes data extraction instruction templates corresponding to the items to be populated.

[0154] Data determination module 110 is used to extract data corresponding to the items to be filled in the disease database from the medical data of disease patients by executing data extraction instructions;

[0155] The data entry module 120 is used to fill data into the items to be filled in the disease database.

[0156] As can be seen from the above description, the disease database filling device proposed in this application can automatically acquire the data corresponding to the items to be filled in the disease database, and automatically fill in the data corresponding to the items to be filled in the disease database, thereby improving the efficiency and accuracy of disease database filling.

[0157] As an optional implementation, another embodiment of this application discloses that the instruction determination module 100 includes: a first determination unit, a second determination unit, and a construction unit;

[0158] The first determining unit is used to determine the data extraction instruction template corresponding to the item to be filled in the disease database from a pre-built medical instruction library according to a preset instruction mapping rule; the instruction mapping rule includes the correspondence between the item to be filled and the data extraction instruction template.

[0159] The second determining unit is used to determine the source of the medical record segment extraction corresponding to the item to be filled in the disease database from the pre-collected medical data of patients with specific diseases, according to the preset medical record segment mapping rules; the medical record segment mapping rules include the correspondence between the item to be filled and the source of the medical record segment extraction.

[0160] The construction unit is used to construct the data extraction instructions corresponding to the items to be filled in the disease database based on the medical record fragment extraction source and the data extraction instruction template corresponding to the items to be filled in the disease database.

[0161] As an optional implementation, another embodiment of this application discloses a construction unit specifically used for:

[0162] The source of the extracted medical record fragments corresponding to the items to be populated in the disease database is filled into the data extraction instruction template corresponding to the items to be populated in the disease database, thus obtaining the data extraction instruction corresponding to the items to be populated in the disease database.

[0163] As an optional implementation, another embodiment of this application discloses a data determination module 110, specifically used for:

[0164] According to the preset medical record fragment mapping rules, extract the medical record fragments corresponding to the items to be filled in the disease database from the medical data of patients with specific diseases;

[0165] By executing the data extraction command, the data corresponding to the items to be filled in the disease database is extracted from the medical record fragments corresponding to the items to be filled in the disease database.

[0166] As an optional implementation, another embodiment of this application discloses that if a disease database item to be filled has a dependency relationship with other disease database items to be filled, then the disease database item to be filled and the other disease database items to be filled that have a dependency relationship with the disease database item to be filled correspond to the same data extraction instruction.

[0167] The data determination module 110 is further used to: extract data corresponding to the items to be filled in the disease database and data corresponding to other items to be filled in the disease database that have a dependency relationship with the items to be filled in the disease database from the medical data of patients with specific diseases by executing data extraction instructions.

[0168] As an optional implementation, another embodiment of this application discloses a data determination module 110, which is further used for:

[0169] Input the data extraction instructions for the items to be populated in the disease database and the medical data of patients with the disease into the pre-built data extraction model to obtain the data to be filled for the items;

[0170] The data extraction model is used to extract data corresponding to the items to be populated in the disease database from the medical data of patients with specific diseases by executing data extraction instructions; the data extraction model is a large language model.

[0171] As an optional implementation, another embodiment of this application discloses that the disease database filling device further includes: a quality inspection module and an output module;

[0172] The quality inspection module is used to inspect the data corresponding to the items to be filled in the disease database according to the pre-set quality inspection rules, and determine the quality inspection information corresponding to the data.

[0173] The data entry module is also used to fill the data into the pending items in the disease database if the quality inspection information indicates that the quality inspection is qualified.

[0174] The output module is used to output a prompt message indicating that the data output is incorrect if the quality inspection information indicates that the quality inspection is unqualified.

[0175] As an optional implementation, another embodiment of this application discloses that the disease database filling device further includes: a training sample acquisition module, an input module, and a parameter adjustment module.

[0176] The training sample acquisition module is used to acquire the training samples and the real data corresponding to the items to be filled in the disease database if the quality inspection information indicates that the quality inspection information is unqualified. The training samples corresponding to the items to be filled in the disease database include: the data extraction instructions and the sample medical data of the items to be filled in the disease database.

[0177] The input module is used to input training samples into the data extraction model to obtain the prediction data corresponding to the training samples.

[0178] The parameter adjustment module is used to adjust the parameters of the data extraction model based on the difference between the real data corresponding to the training samples and the predicted data corresponding to the training samples.

[0179] As an optional implementation, another embodiment of this application discloses that the disease database filling device further includes an identification module.

[0180] The identification module is used to identify the forms to be filled in the disease database and extract at least one item to be filled in the disease database from the forms.

[0181] The disease database filling device provided in this embodiment belongs to the same application concept as the disease database filling method provided in the above embodiments of this application. It can execute the disease database filling method provided in any of the above embodiments of this application and has the corresponding functional modules and beneficial effects for executing the disease database filling method. Technical details not described in detail in this embodiment can be found in the specific processing content of the disease database filling method provided in the above embodiments of this application, and will not be repeated here.

[0182] Exemplary electronic devices

[0183] Another embodiment of this application also provides an electronic device, see [link to relevant documentation] Figure 6 As shown, the device includes:

[0184] Memory 200 and processor 210;

[0185] The memory 200 is connected to the processor 210 and is used to store programs;

[0186] The processor 210 is used to implement the disease database filling method disclosed in any of the above embodiments by running the program stored in the memory 200.

[0187] Specifically, the aforementioned electronic device may also include: a bus, a communication interface 220, an input device 230, and an output device 240.

[0188] The processor 210, memory 200, communication interface 220, input device 230, and output device 240 are interconnected via a bus. Among them:

[0189] A bus can include a pathway for transmitting information between various components of a computer system.

[0190] The processor 210 can be a general-purpose processor, such as a general-purpose central processing unit (CPU), a microprocessor, etc., or an application-specific integrated circuit (ASIC), or one or more integrated circuits used to control the execution of the program of the present invention. It can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0191] Processor 210 may include a main processor, as well as a baseband chip, modem, etc.

[0192] The memory 200 stores a program that executes the technical solution of this invention, and may also store an operating system and other key business functions. Specifically, the program may include program code, which includes computer operation instructions. More specifically, the memory 200 may include read-only memory (ROM), other types of static storage devices capable of storing static information and instructions, random access memory (RAM), other types of dynamic storage devices capable of storing information and instructions, disk storage, flash memory, etc.

[0193] Input device 230 may include a device for receiving user input data and information, such as a keyboard, mouse, camera, scanner, light pen, voice input device, touch screen, pedometer, or gravity sensor.

[0194] Output device 240 may include devices that allow information to be output to a user, such as a display screen, printer, speaker, etc.

[0195] The communication interface 220 may include a device that uses any transceiver to communicate with other devices or communication networks, such as Ethernet, Radio Access Network (RAN), Wireless Local Area Network (WLAN), etc.

[0196] The processor 210 executes the program stored in the memory 200 and calls other devices, which can be used to implement each step of any of the disease database filling methods provided in the above embodiments of this application.

[0197] Exemplary computer program products and storage media

[0198] In addition to the methods and devices described above, embodiments of this application may also be computer program products, which include computer program instructions that, when executed by a processor, cause the processor to perform the steps in the disease database filling methods according to various embodiments of this application as described in the "Exemplary Methods" section of this specification.

[0199] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this application. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0200] Furthermore, embodiments of this application may also be storage media storing a computer program, which is executed by a processor in the steps of the disease database filling method according to various embodiments of this application described in the "Exemplary Methods" section above.

[0201] For the foregoing method embodiments, in order to simplify the description, they are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, because according to this application, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0202] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For apparatus embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0203] The steps in the methods of the various embodiments of this application can be adjusted, merged, or deleted in order according to actual needs, and the technical features described in each embodiment can be replaced or combined.

[0204] The modules and sub-modules in the various embodiments of the present application's devices and terminals can be merged, divided, and deleted according to actual needs.

[0205] It should be understood that the disclosed terminals, devices, and methods can be implemented in other ways, given the several embodiments provided in this application. For example, the terminal embodiments described above are merely illustrative. For instance, the division of modules or sub-modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple sub-modules or modules may be combined or integrated into another module, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or modules, and may be electrical, mechanical, or other forms.

[0206] The modules or submodules described as separate components may or may not be physically separate. The components that constitute a module or submodule may or may not be physical modules or submodules; that is, they may be located in one place or distributed across multiple network modules or submodules. Some or all of the modules or submodules can be selected to achieve the purpose of this embodiment's solution, depending on actual needs.

[0207] Furthermore, the functional modules or sub-modules in the various embodiments of this application can be integrated into one processing module, or each module or sub-module can exist physically separately, or two or more modules or sub-modules can be integrated into one module. The integrated modules or sub-modules described above can be implemented in hardware or in the form of software functional modules or sub-modules.

[0208] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0209] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software unit executed by a processor, or a combination of both. The software unit can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0210] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0211] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for filling in a disease database, characterized in that, include: Based on a pre-built medical instruction library and pre-collected medical data of patients with specific diseases, data extraction instructions corresponding to the items to be populated in the specific disease library are determined; the medical instruction library includes data extraction instruction templates corresponding to the items to be populated; the data extraction instructions corresponding to the items to be populated include the data extraction instruction templates corresponding to the items to be populated and the sources of medical record fragments corresponding to the items to be populated in the medical data of patients with specific diseases. By executing the data extraction instruction, data corresponding to the items to be filled in the disease database is extracted from the medical data of the disease patients. Populate the data into the fields to be populated in the disease database; If the item to be populated in the disease database has a dependency relationship with other items to be populated in the disease database, then the item to be populated in the disease database and the other items to be populated in the disease database that have a dependency relationship with the item to be populated in the disease database correspond to the same data extraction instruction; By executing the data extraction instruction, data corresponding to the items to be populated in the disease database is extracted from the medical data of the disease patients, including: By executing the data extraction instruction, data corresponding to the items to be populated in the disease database and data corresponding to other disease database items that are dependent on the items to be populated in the disease database are extracted from the medical data of the disease patients.

2. The method according to claim 1, characterized in that, Based on a pre-built medical instruction library and pre-collected medical data of patients with specific diseases, data extraction instructions corresponding to the items to be populated in the disease library are determined, including: According to the preset instruction mapping rules, the data extraction instruction template corresponding to the items to be filled in the disease database is determined from the pre-built medical instruction library; the instruction mapping rules include the correspondence between the items to be filled and the data extraction instruction template. According to the preset medical record fragment mapping rules, the source of medical record fragments corresponding to the items to be filled in the disease database is determined from the pre-collected medical data of patients with specific diseases; the medical record fragment mapping rules include the correspondence between the items to be filled and the sources of medical record fragment extraction. Based on the medical record fragment extraction source corresponding to the item to be populated in the disease database and the data extraction instruction template corresponding to the item to be populated in the disease database, a data extraction instruction corresponding to the item to be populated in the disease database is constructed.

3. The method according to claim 2, characterized in that, Based on the medical record fragment extraction source corresponding to the item to be populated in the disease database and the data extraction instruction template corresponding to the item to be populated in the disease database, a data extraction instruction corresponding to the item to be populated in the disease database is constructed, including: The source of the extracted medical record fragments corresponding to the items to be filled in the disease database is filled into the data extraction instruction template corresponding to the items to be filled in the disease database, so as to obtain the data extraction instruction corresponding to the items to be filled in the disease database.

4. The method according to claim 1, characterized in that, By executing the data extraction instruction, data corresponding to the items to be populated in the disease database is extracted from the medical data of the disease patients, including: According to the preset medical record fragment mapping rules, extract the medical record fragments corresponding to the items to be filled in the disease database from the medical data of the disease patients; By executing the data extraction instruction, the data corresponding to the items to be filled in the disease database is extracted from the medical record fragments corresponding to the items to be filled in the disease database.

5. The method according to claim 1, characterized in that, By executing the data extraction instruction, data corresponding to the items to be populated in the disease database is extracted from the medical data of the disease patients, including: The data extraction instructions for the items to be filled in the disease database and the medical data of the patients with the disease are input into the pre-built data extraction model to obtain the data to be filled in the items; The data extraction model is used to extract data corresponding to the items to be filled in the disease database from the medical data of the disease patients by executing the data extraction instructions; wherein, the data extraction model is a large language model.

6. The method according to claim 5, characterized in that, Before populating the data into the disease database items to be populated, the process also includes: According to the pre-set quality inspection rules, the data corresponding to the items to be filled in the disease database is inspected to determine the quality inspection information corresponding to the data. If the quality inspection information indicates that the quality inspection is qualified, then the data will be filled into the items to be filled in the disease database. If the quality inspection information indicates that the quality inspection is unqualified, a prompt message indicating that the data output is incorrect will be output.

7. The method according to claim 6, characterized in that, Also includes: If the quality inspection information indicates that it is unqualified, then the training sample corresponding to the item to be filled in the disease database and the real data corresponding to the training sample are collected; wherein, the training sample corresponding to the item to be filled in the disease database includes: the data extraction instruction corresponding to the item to be filled in the disease database and the sample medical data corresponding to the item to be filled in the disease database. The training samples are input into the data extraction model to obtain the predicted data corresponding to the training samples; Based on the difference between the real data corresponding to the training samples and the predicted data corresponding to the training samples, the parameters of the data extraction model are adjusted.

8. The method according to claim 1, characterized in that, Before determining the data extraction instructions corresponding to the items to be populated in the disease-specific database, based on a pre-built medical instruction library and pre-collected medical data of patients with specific diseases, the process also includes: Identify the forms to be filled in the disease database and extract at least one item to be filled in the disease database from the forms.

9. A disease database filling device, characterized in that, include: The instruction determination module is used to determine the data extraction instruction corresponding to the item to be populated in the disease database based on a pre-built medical instruction library and pre-collected medical data of patients with specific diseases. The medical instruction library includes a data extraction instruction template corresponding to the item to be populated. The data extraction instruction corresponding to the item to be populated includes the data extraction instruction template corresponding to the item to be populated and the source of the medical record fragment corresponding to the item to be populated in the medical data of patients with specific diseases. The data determination module is used to extract the data corresponding to the items to be filled in the disease database from the medical data of the disease patients by executing the data extraction instruction; The data entry module is used to fill the data into the items to be filled in the disease database; If the item to be populated in the disease database has a dependency relationship with other items to be populated in the disease database, then the item to be populated in the disease database and the other items to be populated in the disease database that have a dependency relationship with the item to be populated in the disease database correspond to the same data extraction instruction; The data determination module is further configured to extract, by executing the data extraction instruction, the data corresponding to the items to be filled in the disease database and the data corresponding to other items to be filled in the disease database that have a dependency relationship with the items to be filled in the disease database from the medical data of the disease patients.

10. An electronic device, characterized in that, include: Memory and processor; The memory is connected to the processor and is used to store programs; The processor is configured to implement the disease database filling method as described in any one of claims 1 to 8 by running the program in the memory.

11. A storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the disease database filling method as described in any one of claims 1 to 8.

12. A computer program product, characterized in that, It includes computer program instructions, which, when executed by a processor, cause the processor to implement the disease database filling method as described in any one of claims 1 to 8.

Images