Electronic medical record generation method and system for general nursing

Abstract

The invention discloses an electronic medical record generation method and system for general nursing, relates to the technical field of electric data processing, and can solve the technical problem of how to realize automatic and precise conversion from a paper nursing document to an electronic nursing medical record, and the method comprises the steps: obtaining a historical electronic nursing document, and carrying out the word segmentation processing; determining an importance degree value and a core word segmentation array of each type of electronic nursing documents according to historical electronic nursing documents; clustering the electronic nursing medical record templates according to the importance degree value and the core word segmentation array of each type of electronic nursing documents to obtain a plurality of template clusters; determining an electronic nursing medical record template of the new patient from the plurality of electronic nursing medical record templates according to the paper nursing document of the new patient; and determining the electronic nursing medical record of the new patient according to the electronic nursing document of the new patient and the electronic nursing medical record template of the new patient.

Description

Technical Field

[0001] This invention relates to the field of electronic data processing technology, specifically to a method and system for generating electronic medical records for general nursing care. Background Technology

[0002] With the deepening development of medical informatization, electronic medical record systems have become an important infrastructure for medical institutions to achieve standardized management and information sharing of medical data. In the nursing field, traditional paper-based nursing records need to be converted into standardized electronic nursing medical records to achieve long-term preservation and efficient utilization of medical data. Currently, the common method is for nursing staff to manually select the corresponding electronic template based on the content of the paper document for data entry. This manual operation mode suffers from low efficiency and insufficient accuracy in practical applications, affecting the overall quality of generated electronic nursing medical records. Summary of the Invention

[0003] To address the current technical challenge of automating and accurately converting paper-based nursing documents into electronic nursing records, this invention aims to provide a method and system for generating electronic medical records in general practice nursing. The specific technical solution adopted is as follows: In a first aspect, the present invention provides a method for generating electronic medical records for general nursing care, comprising: acquiring historical electronic nursing documents and performing word segmentation processing; wherein, the historical electronic nursing documents are electronic nursing documents converted from paper nursing documents of historical patients; determining the importance value and core word segmentation array of each type of electronic nursing document based on the historical electronic nursing documents; wherein, the importance value is used to characterize the contribution weight of the corresponding type of electronic nursing document in constituting the electronic nursing medical record, and the core word segmentation array is used to characterize the key information features in the corresponding type of electronic nursing document; clustering electronic nursing medical record templates according to the importance value and core word segmentation array of each type of electronic nursing document to obtain multiple template clusters; wherein, each template cluster contains electronic nursing medical record templates with similar structures and key features; determining the electronic nursing medical record template of a new patient from the multiple electronic nursing medical record templates based on the paper nursing documents of the new patient and the multiple template clusters; and determining the electronic nursing medical record of the new patient based on the electronic nursing documents of the new patient and the electronic nursing medical record template of the new patient.

[0004] In one possible implementation, obtaining historical electronic nursing documents specifically includes: obtaining historical paper nursing documents for patients; wherein, paper nursing documents include various types of documents; converting paper nursing documents into historical electronic nursing documents; and performing word segmentation on each section of the electronic nursing documents to obtain a word segmentation array for each section.

[0005] In one possible implementation, the importance value of each type of electronic nursing document is determined based on historical electronic nursing documents. Specifically, this includes: for each electronic nursing medical record template, obtaining the historical patient set corresponding to the electronic nursing medical record template; obtaining the number of times each type of electronic nursing document appears in the historical patient set, and determining the importance value of each type of electronic nursing document.

[0006] In one possible implementation, based on historical electronic nursing documents, the core word segmentation array for each type of electronic nursing document is determined. Specifically, this includes: for each type of electronic nursing document, calculating the contribution value of each word in each section; wherein the contribution value is used to characterize the frequency of occurrence of the word in the corresponding type of electronic nursing document; selecting a preset number of words as core words from the words in each section in descending order of contribution value; and determining the core word segmentation array for each type of electronic nursing document based on the combination formed by the core words in each section.

[0007] In one possible implementation, electronic nursing record templates are clustered based on the importance value of each type of electronic nursing document and the core word segmentation array. Specifically, this includes: determining the similarity value between each pair of electronic nursing record templates based on the difference in the importance value of the corresponding type of electronic nursing documents and the core word segmentation array; and clustering all electronic nursing record templates according to a preset clustering algorithm and the similarity value between each pair of electronic nursing record templates to obtain multiple template clusters.

[0008] In one possible implementation, based on the new patient's paper nursing documents, the electronic nursing record template for the new patient is determined from multiple electronic nursing record templates. Specifically, this includes: acquiring the new patient's paper nursing documents and performing conversion and word segmentation processing to obtain the new patient's electronic nursing documents; determining the matching degree value between the new patient's electronic nursing documents and each template cluster; identifying the template cluster with the highest matching degree value as the target template cluster, and determining the new patient's electronic nursing record template from the electronic nursing record templates contained in the target template cluster.

[0009] In one possible implementation, the electronic nursing record template for the new patient is determined from the electronic nursing record templates contained in the target template cluster. Specifically, this includes: identifying the difference columns between the various electronic nursing record templates in the target template cluster; detecting whether the electronic nursing document for the new patient contains the difference columns, and obtaining the detection results; and determining the electronic nursing record template for the new patient based on the detection results.

[0010] In one possible implementation, the electronic nursing record template for the new patient is determined based on the test results. Specifically, this includes: if the test results indicate that the electronic nursing document for the new patient does not contain any discrepancy columns, the electronic nursing record template for the new patient is randomly selected from the target template cluster; if the test results indicate that the electronic nursing document for the new patient contains at least one discrepancy column, the electronic nursing record template with the highest number of discrepancy columns is selected as the electronic nursing record template for the new patient.

[0011] In one possible implementation, the electronic nursing record is determined based on the new patient's electronic nursing document and the new patient's electronic nursing record template. Specifically, this includes: extracting information from the electronic nursing document using natural language processing technology; mapping the extracted information to the corresponding columns in the new patient's electronic nursing record template, and generating the new patient's electronic nursing record.

[0012] Secondly, this invention provides an electronic medical record generation system for general practice nursing. The system includes: a document processing module, a feature calculation module, a template clustering module, a template selection module, and a medical record generation module. The document processing module is used to acquire historical electronic nursing documents and perform word segmentation processing; wherein, the historical electronic nursing documents are electronic nursing documents converted from historical patients' paper nursing documents. The feature calculation module is used to determine the importance value and core word segmentation array for each type of electronic nursing document based on the historical electronic nursing documents; wherein, the importance value is used to characterize the contribution of the corresponding type of electronic nursing document in constituting the electronic nursing medical record. The system comprises a weighted core word segmentation array representing key information features in electronic nursing documents of the corresponding type; a template clustering module clustering electronic nursing medical record templates based on the importance value of each type of electronic nursing document and the core word segmentation array, resulting in multiple template clusters; each template cluster contains electronic nursing medical record templates with similar structures and key features; a template selection module determining the electronic nursing medical record template for a new patient from multiple electronic nursing medical record templates based on the new patient's paper nursing document and multiple template clusters; and a medical record generation module determining the electronic nursing medical record for a new patient based on the new patient's electronic nursing document and the new patient's electronic nursing medical record template.

[0013] Thirdly, the present invention provides an electronic device, comprising: a processor and a memory; wherein the memory is used to store one or more programs, the one or more programs including computer-executable instructions, wherein when the electronic device is running, the processor executes the computer-executable instructions stored in the memory to cause the electronic device to perform the electronic medical record generation method for general care as described in the first aspect and any possible implementation thereof.

[0014] Fourthly, the present invention provides a computer-readable storage medium storing one or more programs, the one or more programs including instructions that, when executed by an electronic device of the present invention, cause the electronic device to perform an electronic medical record generation method for general practice nursing as described in the first aspect and any possible implementation thereof.

[0015] Fifthly, the present invention provides a computer program product containing instructions that, when executed on a computer, cause the electronic device of the present invention to perform the electronic medical record generation method for general care as described in the first aspect and any possible implementation thereof.

[0016] In a sixth aspect, the present invention provides a chip system applied to a water meter data acquisition device; the chip system includes one or more interface circuits and one or more processors. The interface circuits and the processors are interconnected via lines; the interface circuits are used to receive signals from the memory of the water meter data acquisition device and send the signals to the processors, the signals including computer instructions stored in the memory. When the processor executes the computer instructions, the water meter data acquisition device performs an electronic medical record generation method for general care as described in the first aspect and any possible design of the present invention.

[0017] The present invention has the following beneficial effects: by automatically analyzing the characteristics of historical electronic nursing documents, a template clustering model based on importance values ​​and core word segmentation arrays is constructed, realizing intelligent template matching and structured filling of paper nursing documents for new patients. This effectively overcomes the problem of inaccurate template selection caused by reliance on human experience in traditional methods, and significantly improves the accuracy, standardization and efficiency of electronic nursing medical record generation. Attached Figure Description

[0018] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention 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 some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the architecture of an electronic medical record generation system for general nursing, provided in one embodiment of the present invention. Figure 2 This is a schematic diagram of the architecture of a document processing module provided in one embodiment of the present invention; Figure 3 This is a schematic diagram of the architecture of a feature calculation module provided in one embodiment of the present invention; Figure 4 This is a flowchart illustrating an electronic medical record generation method for general nursing care, provided as an embodiment of the present invention. Figure 5 This is a flowchart illustrating another method for generating electronic medical records for general nursing care, provided in one embodiment of the present invention. Figure 6 This is a flowchart illustrating another method for generating electronic medical records for general nursing care, provided in one embodiment of the present invention. Figure 7 This is a flowchart illustrating another method for generating electronic medical records for general nursing care, provided as an embodiment of the present invention. Detailed Implementation

[0020] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the specific implementation methods, structures, features, and effects of the present invention will be described in detail below with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0022] For example, such as Figure 1 The diagram shown is an architectural schematic of an electronic medical record generation system for general nursing according to an embodiment of the present invention. The electronic medical record generation system 10 includes: a document processing module 11, a feature calculation module 12, a template clustering module 13, a template selection module 14, and a medical record generation module 15. The modules are described below in sequence: (1) Document processing module 11.

[0023] The document processing module 11 is responsible for converting historical patients' paper-based nursing documents into structured electronic documents and performing word segmentation to provide standardized input data for the subsequent feature calculation module 12.

[0024] Optionally, the document processing module 11 is used to acquire historical electronic nursing documents and perform word segmentation processing. The historical electronic nursing documents are electronic nursing documents converted from historical patients' paper nursing documents.

[0025] For example, such as Figure 2 As shown, the document processing module 11 may include two sub-modules: a document data acquisition sub-module 111 and a conversion and word segmentation sub-module 112.

[0026] (1.1) Document data acquisition submodule 111.

[0027] Optionally, the document data acquisition submodule 111 is used to acquire historical electronic nursing documents.

[0028] Specifically, the document data acquisition submodule 111 connects to the nursing institution's database to batch retrieve historical patients' paper-based nursing documents. These paper-based nursing documents cover various types, including but not limited to personal information documents, nursing record documents, medication record documents, and surgical record documents. If a patient has undergone multiple surgeries or taken multiple medications, multiple paper-based nursing documents of the same type will be retrieved.

[0029] (1.2) Transformation word segmentation submodule 112.

[0030] Optionally, the word segmentation submodule 112 is used to perform word segmentation processing on historical electronic nursing documents.

[0031] For example, before performing word segmentation, the forwarding word segmentation submodule 112 uses optical character recognition (OCR) technology to scan and recognize characters in the paper nursing documents obtained by the paper document acquisition submodule 111, converting the handwritten or printed text in the paper documents into editable electronic text, and finally generating electronic nursing documents for historical patients.

[0032] Furthermore, the word segmentation submodule 112 performs word segmentation processing on the electronic nursing documents generated by the OCR conversion submodule 112 according to the "column" dimension—that is, it calls the word segmentation tool to segment the text content of each preset column in the electronic nursing document, and obtains the word segmentation array for each column. For example, if the word segmentation tool is Jieba, the column content "body temperature 36.5℃" will be segmented into [body temperature, 36.5℃].

[0033] (2) Feature calculation module 12.

[0034] The feature calculation module 12 is responsible for classifying the types of electronic nursing documents based on the historical electronic nursing documents and word segmentation arrays output by the document processing module 11, and calculating the importance value and core word segmentation array of each type of electronic nursing document, so as to provide the template clustering module 13 with core feature parameters for template clustering.

[0035] Optionally, the feature calculation module 12 is used to determine the importance value and core word segmentation array for each type of electronic nursing document based on historical electronic nursing documents. The importance value represents the contribution weight of the corresponding type of electronic nursing document in constituting the electronic nursing medical record, and the core word segmentation array represents the key information features in the corresponding type of electronic nursing document.

[0036] For example, such as Figure 3 As shown, the feature calculation module 12 may include three sub-modules: document type classification sub-module 121, importance value calculation sub-module 122, and core word segmentation array determination sub-module 123.

[0037] (2.1) Document type division sub-module 121.

[0038] Specifically, the document type classification submodule 121 is used to classify the historical electronic nursing documents output by the document processing module 11 before determining the importance value and core word segmentation array of each type of electronic nursing document. For example, by analyzing the similarity of historical electronic nursing documents, documents that "record patient basic information and allergy history" are uniformly classified as "personal information document type", and documents that "record vital signs and nursing operations" are uniformly classified as "nursing record document type", ultimately determining the type set of all electronic nursing documents in the system.

[0039] (2.2) Importance value calculation submodule 122.

[0040] Optionally, the importance value calculation submodule 122 is used to determine the importance value of each type of electronic nursing document, specifically including: for each electronic nursing medical record template, obtaining the historical patient set corresponding to the electronic nursing medical record template; then, obtaining the number of times each type of electronic nursing document appears in the historical patient set, and determining the importance value of each type of electronic nursing document.

[0041] (2.3) Core word segmentation array determines submodule 123.

[0042] Optionally, the core word segmentation array determination submodule 123 is used to determine the core word segmentation array for each type of electronic nursing document. Specifically, this includes: first, for each type of electronic nursing document, calculating the contribution value of each word in each section; then, selecting a preset number of words with the highest contribution values ​​from the words in each section as core words; finally, determining the core word segmentation array for each type of electronic nursing document based on the combinations formed by the core words in each section. The contribution value is used to characterize the frequency of word occurrence in the corresponding type of electronic nursing document.

[0043] (3) Template clustering module 13.

[0044] The template clustering module 13 is responsible for intelligently grouping all electronic nursing medical record templates. Based on the importance values ​​of various document types and the core word segmentation array output by the feature calculation module 12, this module automatically merges electronic nursing medical record templates with similar structures and key features into multiple template clusters by quantitatively analyzing the similarity between templates, thus establishing a classification basis for subsequent template matching.

[0045] Optionally, the template clustering module 13 is used to cluster electronic nursing record templates based on the importance value of each type of electronic nursing document and the core word segmentation array, resulting in multiple template clusters. Each template cluster contains electronic nursing record templates with similar structures and key features.

[0046] Specifically, the template clustering module 13 first calculates the overall similarity value between every two electronic nursing record templates. For any two templates, the system compares their importance values ​​for each document type and uses the Jaccard correlation coefficient algorithm to calculate the similarity of the corresponding core word segmentation arrays. By weighting and integrating these dimensional indicators, a similarity quantification value between 0 and 1 is finally generated. The larger the value, the closer the overall structure and content features of the two templates are.

[0047] After determining the similarity relationships between templates, the template clustering module 13 automatically analyzes the optimal number of clusters using the silhouette coefficient method. This method evaluates the clustering effect under different cluster numbers to determine the total number of template clusters that achieves the optimal balance between intra-cluster compactness and inter-cluster separation, providing key parameters for subsequent clustering operations.

[0048] Finally, the template clustering module 13 uses the calculated template similarity values ​​as a distance metric and applies the Spectral Clustering algorithm to group all electronic nursing record templates. This process generates a specified number of template clusters, where the electronic nursing record templates within each cluster maintain a high degree of consistency in document structure, importance distribution, and core word segmentation features. These template clusters are then output to the template selection module 14 as the basis for accurately matching electronic nursing record templates to new patients.

[0049] (4) Template selection module 14.

[0050] The template selection module 14 is responsible for accurately selecting the most suitable electronic nursing record templates for new patients. This module processes the paper nursing documents of new patients and combines them with the template cluster information output by the template clustering module 13 to complete the entire process of electronic document conversion, cluster-level matching degree calculation, target cluster determination, and final template selection, providing an accurate template basis for the medical record generation module 15.

[0051] Optionally, the template selection module 14 is used to determine the electronic nursing record template for a new patient from multiple electronic nursing record templates based on the new patient's paper nursing documents and multiple template clusters.

[0052] Specifically, the template selection module 14 first acquires the paper nursing documents of new patients, converts them into electronic nursing documents using optical character recognition technology, and then performs word segmentation processing to obtain word segmentation arrays for each section. Based on this, the module further generates core word segmentation arrays for each type of electronic nursing document for new patients, providing feature-based support for subsequent matching calculations.

[0053] Following this, the template selection module 14 calculates the matching degree value between the new patient's electronic nursing document and each template cluster. This calculation process comprehensively considers the distribution of importance values ​​of each electronic nursing medical record template within the template cluster and its core word segmentation array features. By weighted fusion of similarity indicators of different types of documents, a quantitative value reflecting the degree of matching between the new patient and each template cluster is finally generated.

[0054] Furthermore, after obtaining the matching degree values ​​of all template clusters, the template selection module 14 compares these values ​​and identifies the template cluster with the highest matching degree value as the target template cluster. This step ensures that the selected template set best matches the current new patient's nursing needs in terms of structure and content characteristics.

[0055] Finally, the template selection module 14 filters the final templates within the target template cluster. The module analyzes the differences in columns between templates within the cluster and checks whether the new patient's electronic nursing document contains these differing columns. Based on the detection results, it adopts an appropriate filtering strategy: if no differing columns are found, a template is randomly selected; if differing columns are found, the template with the highest matching degree to the patient's document is selected. The final determined electronic nursing record template will be output to the medical record generation module 15 along with the new patient's electronic nursing document.

[0056] (5) Medical record generation module 15.

[0057] The medical record generation module 15 is responsible for combining processed patient information with a matching template to generate standardized electronic nursing medical records. This module receives the final electronic nursing medical record template from the template selection module 14 and the corresponding new patient electronic nursing document, and uses natural language processing technology to intelligently extract and structure information, ultimately outputting a complete electronic nursing medical record that conforms to medical standards.

[0058] For example, the medical record generation module 15 first uses natural language processing technology to perform in-depth analysis and extract key information from the new patient's electronic nursing document. This process can accurately identify and extract various medical data contained in the document, including but not limited to vital sign data, medication details, surgical information, and other key clinical elements, providing structured data support for subsequent template filling.

[0059] Next, the medical record generation module 15 maps the extracted key information to the corresponding columns of the electronic nursing medical record template according to the field correspondence. By establishing field mapping rules between the source document and the target template, it ensures that various types of medical information can be accurately filled into the designated locations in the template. For columns that exist in the template but for which the source document does not provide corresponding information, the system will automatically mark them with specific identifiers to maintain the structural integrity of the document, ultimately generating electronic nursing medical records that conform to medical standards and are accurate in content.

[0060] Finally, the medical record generation module 15 outputs the final electronic nursing medical record to the medical institution's electronic medical record management system. This output interface ensures that newly effective nursing medical records can be promptly incorporated into the medical institution's information management system, allowing medical staff to view, modify, and archive them, thus completing the entire transformation process from paper documents to standardized electronic nursing medical records.

[0061] The above describes the electronic medical record generation system 10 and its included modules.

[0062] For example, such as Figure 4 The diagram shown is a flowchart illustrating a method for generating electronic medical records for general nursing care according to an embodiment of the present invention, including the following steps: S401. Obtain historical electronic nursing documents and perform word segmentation.

[0063] Among them, historical electronic nursing documents are electronic nursing documents converted from historical paper nursing documents of patients.

[0064] In one possible implementation, S401 can be executed by the document processing module 11 described above. First, it acquires historical paper-based nursing documents covering multiple types; then, it converts the paper documents into editable historical electronic nursing documents; finally, it performs word segmentation processing on the electronic nursing documents according to column dimensions, generating a word segmentation array that reflects the key information of the document. Specifically, it includes the following steps: (1) Obtain paper nursing documents of historical patients.

[0065] Paper-based nursing documents include various types of documents.

[0066] For example, this sub-step can be executed by the document data acquisition submodule 111 in the document processing module 11, specifically including: the document data acquisition submodule 111 acquiring historical patients' paper-based nursing document data by connecting to the medical institution's medical record management system. The aforementioned document types cover personal information documents, nursing record documents, medication record documents, and surgical record documents.

[0067] Optionally, to ensure data integrity, when a patient has multiple medical records, the document data acquisition submodule 111 can acquire multiple corresponding paper documents of the same type. Furthermore, the document data acquisition submodule 111 records the patient identifier associated with each paper document and its corresponding electronic nursing record template identifier to ensure data traceability in subsequent processing.

[0068] (2) Convert paper nursing documents into historical electronic nursing documents.

[0069] For example, this sub-step can be performed by the conversion and word segmentation sub-module 112 in the document processing module 11, specifically including: the conversion and word segmentation sub-module 112 uses optical character recognition technology to convert the paper nursing document obtained by the document data acquisition sub-module 111 into an electronic nursing document.

[0070] During the conversion process, the conversion word segmentation submodule 112 first scans the paper document into an image file and optimizes the image quality through image preprocessing technology. Then, it identifies the text content in the image and converts it into an editable text format, ultimately generating an electronic nursing document that is consistent with the content of the original paper document.

[0071] (3) Perform word segmentation on each section of the electronic nursing document to obtain a word segmentation array for each section.

[0072] For example, this sub-step can be executed by the conversion and word segmentation sub-module 112 in the document processing module 11. Specifically, the conversion and word segmentation sub-module 112 divides the converted electronic nursing documents according to the column dimension, and uses word segmentation technology to segment the text content of each column, generating a corresponding word segmentation array. The conversion and word segmentation sub-module 112 also records the column, document type, and associated patient information of each word segment, providing a data foundation for the construction of the core word segmentation array in the subsequent feature calculation module.

[0073] Thus, the document processing module 11 ensures that the data format of historical electronic nursing documents is unified and can be directly used for template feature analysis, ensuring that the subsequent feature calculation module 12 can perform accurate importance value and core word segmentation array calculations based on standardized data.

[0074] S402. Based on historical electronic nursing documents, determine the importance value and core word segmentation array for each type of electronic nursing document.

[0075] Among them, the importance value is used to characterize the contribution weight of the corresponding type of electronic nursing document in the formation of electronic nursing medical record, and the core word segmentation array is used to characterize the key information features in the corresponding type of electronic nursing document.

[0076] For example, S402 can be executed by the feature calculation module 12 in the aforementioned electronic medical record generation system 10. The process by which the feature calculation module 12 determines the aforementioned two parameters is described below: (1) Classify historical electronic nursing documents.

[0077] For example, this sub-step can be performed by the document type segmentation sub-module 121 in the feature calculation module 12.

[0078] Specifically, before determining the importance value and core word segmentation array for each type of electronic nursing document, the document type classification submodule 121 classifies the historical electronic nursing documents output by the document processing module 11. For example, by analyzing the similarity of historical electronic nursing documents, documents that "record patient basic information and allergy history" are uniformly classified as "personal information document type," and documents that "record vital signs and nursing operations" are uniformly classified as "nursing record document type," ultimately determining the type set of all electronic nursing documents in the system.

[0079] (2) Determine the importance value of each type of electronic nursing document based on historical electronic nursing documents.

[0080] For example, this sub-step can be executed by the importance value calculation sub-module 122 in the feature calculation module 12, and may specifically include the following sub-steps: First, for each electronic nursing record template, the importance value calculation sub-module 122 obtains the historical patient set corresponding to the electronic nursing record template; then, the importance value calculation sub-module 122 obtains the number of times each type of electronic nursing document appears in the historical patient set, and determines the importance value of each type of electronic nursing document accordingly. It should be noted that the specific process of the importance value calculation sub-module 122 determining the importance value of each type of electronic nursing document according to the aforementioned two sub-steps is described in S501-S502 below, and will not be repeated here.

[0081] It is understandable that the aforementioned electronic nursing record template refers to a pre-designed document framework for general nursing scenarios, containing a fixed column structure and information specifications, used to unify the format and content dimensions of electronic nursing records. Specifically, while including multiple pre-designed columns, the electronic nursing record template also supports adding, deleting, or adjusting columns according to nursing needs. For example, the template is mainly constructed in three ways: extracting common columns based on the basic standards for nursing record writing published by authoritative institutions; directly importing standardized templates currently in use in the medical institution's existing information system; or, through structural analysis of a large number of historical electronic nursing documents, clustering high-frequency column sets and solidifying them into a template.

[0082] In another possible implementation, the importance value calculation submodule 122 can also use a time-weighted frequency statistical method to determine the importance value of each type of electronic nursing document. This method assigns different time weights to historical documents from different periods when counting the frequency of document types, typically assigning higher weights to more recent documents. This weighted calculation of importance values ​​helps to better reflect the representativeness of recent documents in the context of continuously updated nursing standards.

[0083] In another possible implementation, the importance value calculation submodule 122 can also integrate domain expert knowledge and objective data when determining the importance value of each type of electronic nursing document. By combining the theoretical importance scores of nursing experts for various types of documents with their frequency of occurrence in actual historical data, the importance value is formed after weighted calculation. This method is particularly suitable for assessing new types of documents with insufficient historical data.

[0084] Therefore, the importance value calculation submodule 122 quantifies the contribution weight of document types to the electronic nursing records by statistically analyzing the frequency of occurrence of document types in the historical patient set, and completes feature extraction from the overall contribution dimension of document types, ensuring that the subsequent template clustering module 13 can carry out clustering analysis based on clear feature parameters.

[0085] (3) Based on historical electronic nursing documents, determine the core word segmentation array for each type of electronic nursing document.

[0086] For example, this sub-step can be executed by the core word segmentation array determination sub-module 123 in the feature calculation module 12, specifically including the following sub-steps: First, for each type of electronic nursing document, the core word segmentation array determination sub-module 123 calculates the contribution value of each word in each section; then, the core word segmentation array determination sub-module 123 selects a preset number of words with the highest contribution values ​​from the words in each section as core words; finally, the core word segmentation array determination sub-module 123 determines the core word segmentation array for each type of electronic nursing document based on the combination formed by the core words in each section. The contribution value is used to characterize the frequency of word occurrence in the corresponding type of electronic nursing document. It should be noted that the specific process of the core word segmentation array determination sub-module 123 determining the importance value of each type of electronic nursing document according to the aforementioned three sub-steps is described in S503-S505 below, and will not be repeated here.

[0087] In another possible implementation, the core word segmentation array determination submodule 123 can also use an evaluation method based on word frequency-inverse document frequency weighting when determining the core word segmentation array. This method calculates the importance index by comprehensively considering the frequency of word segmentation in specific types of documents and its distribution in all documents, which can effectively reduce the weight of general vocabulary, thereby highlighting type-specific keywords more prominently.

[0088] In another possible implementation, the core word segmentation array determination submodule 123 can also use mutual information measurement to filter core words by calculating the correlation strength between word segments and specific document types when determining the core word segmentation array. Those word segments with the highest correlation with the type will be selected into the core array. This method can significantly enhance the ability of the core word segmentation array to distinguish between different types of documents.

[0089] Therefore, the core word segmentation array determination submodule 123 calculates the representativeness (i.e. contribution value) of word segmentation in document type, selects core word segmentation that can reflect the key features of document and forms an array, completes feature extraction from the key information dimension of document content, and ensures that the subsequent template clustering module 13 can carry out clustering analysis based on clear feature parameters.

[0090] S403. Cluster the electronic nursing record templates according to the importance value of each type of electronic nursing document and the core word segmentation array to obtain multiple template clusters.

[0091] Each of the multiple template clusters contains electronic nursing record templates with similar structures and key features.

[0092] For example, S403 can be executed by the template clustering module 13 described above, specifically including the following sub-steps: First, the template clustering module 13 determines the similarity value between each pair of electronic nursing medical record templates based on the difference in importance values ​​of corresponding type electronic nursing documents and the core word segmentation array; then, the template clustering module 13 clusters all electronic nursing medical record templates according to a preset clustering algorithm and the similarity value between each pair of electronic nursing medical record templates, obtaining multiple template clusters. It should be noted that the specific process of the template clustering module 13 clustering all electronic nursing medical record templates according to the aforementioned two sub-steps is described in S601-S602 below, and will not be repeated here.

[0093] In another possible implementation, the template clustering module 13 can also employ a similarity calculation method based on a vector space model during clustering. This method maps the core word segmentation arrays of each template to word vectors in a high-dimensional space, quantifies the similarity of core word segmentation features by calculating the cosine similarity between these vectors, and then performs weighted fusion by combining the differences in importance values ​​of different document types to finally obtain the overall similarity measure between templates. This method is particularly suitable for processing high-dimensional data scenarios with complex word segmentation features.

[0094] Furthermore, in terms of clustering algorithm selection, the template clustering module 13 can also adopt a hierarchical clustering algorithm as an alternative. This algorithm constructs a hierarchical clustering tree in a bottom-up or top-down manner, without needing to pre-specify the number of clusters, and can automatically discover the hierarchical relationships between templates. It is particularly suitable for application scenarios where the differences in template features are not obvious, while preserving the inherent hierarchical structure between templates.

[0095] Therefore, the template clustering module 13 quantifies the similarity between electronic nursing medical record templates, and classifies templates with similar structures and key features into one category (i.e., "template cluster"), providing a classification basis for the subsequent template selection module 14 to quickly match the target template cluster of new patients.

[0096] S404. Based on the new patient's paper nursing documents and multiple template clusters, determine the new patient's electronic nursing record template from multiple electronic nursing record templates.

[0097] For example, S404 can be executed by the template selection module 14 described above, specifically including the following sub-steps: First, the template selection module 14 obtains the paper nursing document of the new patient and performs conversion and word segmentation processing to obtain the electronic nursing document of the new patient; then, the template selection module 14 determines the matching degree value between the electronic nursing document of the new patient and each template cluster; finally, the template selection module 14 determines the template cluster with the highest matching degree value as the target template cluster, and determines the electronic nursing medical record template of the new patient from the electronic nursing medical record templates contained in the target template cluster. It should be noted that the specific process of the template selection module 14 determining the electronic nursing medical record template of the new patient according to the aforementioned three sub-steps can be found in S701-S703 below, and will not be repeated here.

[0098] In another possible implementation, when the template selection module 14 determines the electronic nursing record template for a new patient, a multi-cluster fusion selection mechanism can also be used. This mechanism not only considers a single best-matching cluster, but also comprehensively evaluates the features of multiple template clusters with the highest matching degree. By extracting the common features of these clusters, a fused template feature is constructed, and then the most suitable individual is selected from all templates. This method can effectively handle new patient cases with complex features and avoid matching bias caused by the limitations of single-cluster selection.

[0099] Furthermore, the template selection module 14 can also employ a human-machine collaborative decision-making model. The template selection module 14 automatically calculates and recommends candidate template clusters with high matching degrees, and medical staff make the final selection based on clinical experience and the patient's specific circumstances. This implementation combines the computational power of artificial intelligence with the clinical judgment of professionals, providing more reliable matching results when dealing with ambiguous features or special cases.

[0100] Therefore, the template selection module 14 first converts the new patient's paper nursing documents into standardized electronic documents and segments them to generate feature data that can be used for matching; then it calculates the matching degree between the new patient data and each template cluster to quantify the degree of fit; finally, it selects the template cluster with the highest matching degree as the target cluster and determines the new patient's electronic nursing record template from this cluster. This solves the problems of a large number of templates and the ease of error in manual selection in existing technologies. By first clustering and filtering and then precise matching, the efficiency and accuracy of template selection are greatly improved.

[0101] S405. Based on the new patient's electronic nursing documents and the new patient's electronic nursing record template, determine the new patient's electronic nursing record.

[0102] For example, S405 can be executed by the medical record generation module 15 described above, specifically including the following steps: (1) Extract information from electronic nursing documents using natural language processing technology.

[0103] Optionally, the medical record generation module 15 can use Neuro-Linguistic Programming (NLP) technology to extract structured information from the electronic nursing documents of the new patient output by the template selection module 14, and then output a structured information extraction result set.

[0104] (2) Map the extracted information to the corresponding columns of the new patient's electronic nursing record template to generate the new patient's electronic nursing record.

[0105] Optionally, the mapping logic in this step is as follows: using the "electronic nursing record template for new patients" determined by template selection module 14 as a framework, this template contains a preset column system. Mapping is achieved by matching information category tags with column names, specifically including direct matching, semantic matching, multi-value merging, etc. If a certain column of the template does not have corresponding extracted information, for example, the template contains a "postoperative complications" column, but the new patient has not undergone surgery and there is no relevant information in the electronic nursing document, then the column is marked "no relevant record" to ensure the integrity of the medical record structure.

[0106] Thus, after mapping and filling, the template selection module 14 forms a new patient's electronic nursing record containing all template columns and with complete content. This record not only meets the structured requirements of the template but also fully preserves the new patient's personalized nursing information, and can be directly used for clinical nursing records, archiving, and subsequent diagnosis and treatment reference.

[0107] Based on the above technical solution, this embodiment of the invention automatically analyzes the characteristics of historical electronic nursing documents and constructs a template clustering model based on importance values ​​and core word segmentation arrays. This enables intelligent template matching and structured filling of paper nursing documents for new patients, effectively overcoming the problem of inaccurate template selection caused by reliance on human experience in traditional methods, and significantly improving the accuracy, standardization and efficiency of electronic nursing medical record generation.

[0108] For example, in combination Figure 4 ,like Figure 5 The diagram shown is a flowchart illustrating another method for generating electronic medical records for general nursing care according to an embodiment of the present invention. In this method, based on historical electronic nursing documents, the importance value and core word segmentation array for each type of electronic nursing document are determined, specifically including the following steps: S501. For each electronic nursing record template, obtain the set of historical patients corresponding to the electronic nursing record template.

[0109] For example, the importance value calculation submodule 122 selects "historical patients who have generated electronic nursing records using the nth template" from the historical electronic nursing documents obtained by the document processing module 11 for each electronic nursing medical record template in the system, forming a set of historical patients corresponding to the template, denoted as containing M patients (M is a positive integer, M≥1).

[0110] S502. Obtain the number of times each type of electronic nursing document appears in the historical patient set, and determine the importance value of each type of electronic nursing document.

[0111] In this step, the importance value calculation submodule 122 calculates the total number of times each type of electronic nursing document (let's call it the r-th type, r=1, 2, ..., R; R is the total number of document types) that the document processing module 11 has divided appears in the historical electronic nursing documents of the above M patients. This number is denoted as sum(r, M). For example, if the "nursing record document type" appears a total of 120 times in the historical documents of 50 patients, then sum(r, M) = 120.

[0112] Furthermore, the importance value calculation submodule 122 calculates the importance value for each type of electronic nursing document using the following formula: In the formula, This represents the importance value of the r-th type of nursing document. This indicates the total number of patients. This represents the number of times the r-th type of nursing document appears in all M patients. The more times it appears, the higher the importance value of the r-th type of nursing document.

[0113] S503. For each type of electronic nursing document, calculate the contribution value of each word segment in each column. The contribution value is used to characterize the frequency of word occurrence in the corresponding type of electronic nursing document.

[0114] Specifically, the core word segmentation array determines the contribution value of each word segmentation for each category using the following formula: In the formula, This represents the contribution value of the i-th word in a certain section of a nursing document of type r. This represents the total number of patients. m represents the patient's serial number. This indicates the number of times the document of type r appears in the nursing documentation for the m-th patient. h represents the sequence number of a single document. This represents the number of times the i-th word appears in the h-th document within the r-th type of document for the m-th patient. This represents the total number of words in the h-th document within the r-th type of document for the m-th patient.

[0115] It should be noted that since the nursing documents of the rth type for the mth patient may appear multiple times, calculating the frequency of the ith word segment might be too high. Therefore, a three-tiered logic is used to quantify this: the word segmentation ratio in a single document, the cumulative ratio across multiple documents, and the average ratio at the patient level. Specifically, the frequency of the ith word segmentation in the hth document is used as the starting point. Divide by the total number of words in the document First, the absolute number of word segmentations is converted into a relative proportion to avoid inflated numbers for single long documents; second, for all documents of this type for the same patient (total... The word segmentation percentages of each document are accumulated over time, preserving the positive association that more frequent occurrences of a word indicate stronger core meaning, while avoiding the dilution of individual contributions across multiple documents. Finally, the cumulative percentages of all M related patients are averaged to eliminate the impact of differences in document quantity among patients, ultimately resulting in... The calculation not only solves the problem of excessively high original frequency, but also better meets the requirement of word segmentation to be representative of the core document type.

[0116] S504. Select a preset number of words as core words from the word segments of each column, sorted in descending order of contribution value.

[0117] Optionally, the preset quantity can be set to an empirical value of 15. This empirical value is sufficient for moderately complex electronic nursing document types to cover the core information of the section while avoiding redundancy. Furthermore, the preset quantity can be reduced for low-complexity electronic nursing document types and increased for high-complexity electronic nursing document types.

[0118] For example, the preset number of values ​​is determined by accurately covering the key information in electronic nursing documents. Based on this, a baseline range is defined according to document type complexity: 8-12 for low-complexity documents, 13-17 for medium-complexity documents, and 18-22 for high-complexity documents. Simultaneously, adjustments are made based on the clinical importance of each section. For example, key sections like "vital signs" are increased by 2-3 from the baseline range, while auxiliary sections like "signature" are reduced by 2-3. This balances value stability with information coverage flexibility, ensuring that the core word segmentation array accurately represents the key features of the corresponding document type.

[0119] S505. Based on the combination formed by the core word segmentation of each column, determine the core word segmentation array for each type of electronic nursing document.

[0120] For example, the core word segmentation array determination submodule 123 combines the core words of all columns of the r-th type of electronic nursing document according to the column order to form the core word segmentation array of this type of document, denoted as . Therefore, by performing S505 on each type of electronic nursing document in each template, the core word segmentation array of all types of electronic nursing documents in all templates can be obtained.

[0121] Based on the above technical solution, this invention achieves automated extraction and standardized representation of electronic nursing record template features by objectively quantifying historical nursing document data. Specifically, the calculation of importance values ​​based on the frequency of document types accurately reflects the actual contribution weight of different types of nursing documents in the composition of the medical record; while the construction of a core word segmentation array combined with word segmentation frequency analysis effectively extracts key information features from various types of documents. This data-driven feature extraction method effectively overcomes the problem of inaccurate template matching caused by reliance on manual experience in traditional methods, laying a reliable data foundation for subsequent accurate template clustering and matching, thereby significantly improving the accuracy and consistency of electronic nursing record generation.

[0122] For example, in combination Figure 4 ,like Figure 6 The diagram shown is a flowchart illustrating another method for generating electronic medical records for general nursing care, provided by an embodiment of the present invention. In this method, electronic nursing medical record templates are clustered based on the importance value of each type of electronic nursing document and the core word segmentation array. Specifically, the method includes the following steps: S601. Based on the difference in importance values ​​of corresponding electronic nursing documents of each pair of electronic nursing medical record templates and the core word segmentation array, determine the similarity value between each pair of electronic nursing medical record templates.

[0123] For example, the template clustering module 13 can calculate the similarity value between every two electronic nursing record templates using the following formula: In the formula, This represents the similarity value between the p-th electronic nursing record template and the q-th electronic nursing record template. This indicates the total number of all types of electronic nursing records. This represents the importance value of the r-th type of nursing document in the p-th electronic nursing record. This represents the importance value of the r-th type of nursing document in the q-th electronic nursing record. The "+1" in the denominator of the formula is to avoid the denominator being 0. This represents the Jaccard correlation coefficient. This represents the core word segmentation array of the r-th type of nursing document in the p-th electronic nursing medical record template. This represents the core word segmentation array of the r-th type of nursing document in the q-th electronic nursing medical record template.

[0124] It should be noted that, The smaller the value, the more similar the importance of the r-th electronic nursing record template among the p-th and q-th electronic nursing record templates. Meanwhile... The larger the value, the more similar the core words of the p-th and q-th electronic nursing medical record templates are in the r-th electronic nursing medical record template. The larger the similarity value, the greater the similarity between the p-th and q-th electronic nursing medical record templates.

[0125] S602. Based on the preset clustering algorithm and the similarity value between every two electronic nursing medical record templates, cluster all electronic nursing medical record templates to obtain multiple template clusters.

[0126] Optionally, the preset clustering algorithm may be the spectral clustering algorithm.

[0127] For example, the template clustering module 13 collects the similarity values ​​of all electronic nursing record templates (total number N) in the system, forming an N×N similarity matrix, where the element in the p-th row and q-th column of the matrix is... Furthermore, the template clustering module 13 uses a preset clustering algorithm to cluster the similarity matrix—by setting the number of clusters U, the N templates are divided into U template clusters. It should be noted that the aforementioned number of clusters can be determined based on the silhouette coefficient method to determine the optimal number of clusters U.

[0128] Based on the above technical solution, this embodiment of the invention calculates the similarity value between any two templates based on the differences in importance values ​​of electronic nursing documents corresponding to different template types and the core word segmentation array; then, it groups all templates according to their similarity values ​​using a preset clustering algorithm, ultimately obtaining multiple template clusters. This process solves the problem of the large number of templates and the difficulty in efficient matching in the prior art. By categorizing templates by features through clustering, subsequent selection of templates for new patients can focus on the most similar template clusters, improving matching efficiency and accuracy.

[0129] For example, in combination Figure 4 ,like Figure 7 The diagram shown is a flowchart illustrating another method for generating electronic medical records for general nursing care according to an embodiment of the present invention. In this method, based on the new patient's paper nursing documents and multiple template clusters, an electronic nursing medical record template for the new patient is determined from multiple electronic nursing medical record templates. Specifically, this includes the following steps: S701. Obtain the paper nursing documents of the new patient and perform conversion and word segmentation to obtain the electronic nursing documents of the new patient.

[0130] For example, in this step, the template selection module 14 reuses the technical process of the document processing module 11 to convert the unstructured paper nursing documents of new patients into structured electronic documents and word segmentation arrays, providing standardized input for subsequent matching degree calculation.

[0131] S702. Determine the matching value between the electronic nursing documents of new patients and each template cluster.

[0132] For example, the template selection module 14 calculates the matching degree value between the new patient's electronic nursing document and each template cluster using the following formula: In the formula, This represents the degree of match between the new patient's electronic nursing care document and the electronic nursing record template in cluster u. T represents the number of electronic nursing care templates in cluster u. R represents the total number of all types of electronic nursing records. (Calculation method is the same) ) represents the importance value of the nursing document of type r in the t-th electronic nursing record model of cluster u. This represents the Jaccard correlation coefficient. The core word segmentation array representing the r-th type of electronic nursing document for a new patient. This represents the core word segmentation array of the r-th type of electronic nursing document in the t-th electronic nursing medical record template of the u-th cluster.

[0133] It should be noted that, A larger value indicates a higher importance of the core word segmentation in the r-th type of electronic nursing document within the t-th electronic nursing record template in the u-th cluster. This results in a higher similarity weight between the core word segmentation array of the new patient and the core word segmentation array of the n-th electronic nursing record template during template matching. Since the denominators in the above formula are all parameters used in practical applications, none are zero, ensuring that a zero denominator will not render the formula calculation meaningless.

[0134] S703. The template cluster with the highest matching degree value is determined as the target template cluster, and the electronic nursing record template of the new patient is determined from the electronic nursing record templates contained in the target template cluster.

[0135] In this step, the most suitable target template cluster is located by sorting the matching degree values, and then the specific template that best matches the characteristics of the new patient document is selected from this cluster.

[0136] Specifically, the template selection module 14 sorts all U template clusters in descending order of their matching scores and selects the template cluster with the highest matching score as the target template cluster. Then, from the T electronic nursing record templates contained in the target template cluster, and considering the column characteristics of the new patient's electronic nursing document, the final new patient's electronic nursing record template is determined, including the following steps: (1) Determine the differences between the various electronic nursing record templates in the target template cluster.

[0137] Among them, the "differentiated columns" refer to columns that are included in some electronic nursing record templates within the target template cluster but not in others; that is, "non-general columns" that are not common to all templates. For example, if the target template cluster is the "Pediatric Nursing Template Cluster" and contains three templates: Template A contains the "Weight Gain Record" column, Template B contains the "Weight Gain Record" and "Complementary Food Introduction Record" columns, and Template C does not contain the above two columns, then "Weight Gain Record" and "Complementary Food Introduction Record" are both differentiated columns of this cluster.

[0138] Optionally, the template selection module 14 extracts the complete column list of each electronic nursing medical record template within the target template cluster. By comparing each column in the complete column list, it filters out columns that appear less than T times, i.e. columns not included in all templates, and summarizes them to form a set of differential columns for the target template cluster.

[0139] (2) Check whether the electronic nursing documents of new patients contain the difference column and obtain the test results.

[0140] Optionally, the template selection module 14 extracts all the columns contained in the new patient's electronic nursing document to form a new patient column list; then, it matches the new patient column list with the set of differences in the target template cluster to determine whether each difference column in the set of differences exists in the new patient column list.

[0141] For example, the detection result is represented in terms of included state, specifically: If none of the difference categories in the difference category set appear in the new patient category list, it is indicated as "the electronic nursing document for the new patient does not contain any difference categories". If at least one of the difference categories appears in the new patient category list, it is indicated as "The new patient's electronic care document contains at least one difference category".

[0142] (3) Based on the test results, determine the electronic nursing record template for new patients.

[0143] In this step, the test results directly determine the template selection strategy: if the new patient document does not contain any difference columns, it means that the difference columns are not of practical significance to the patient, and a template can be selected through a simplification strategy; if it contains at least one difference column, it means that the difference columns are related to the patient's care needs, and a template should be selected through a precise matching strategy.

[0144] For example, the exact matching strategy is set as follows: a. If the electronic nursing record template for the new patient does not contain a difference column in the test results characterizing the new patient, randomly select the electronic nursing record template for the new patient from the target template cluster.

[0145] When the test results indicate that the new patient document does not contain any difference columns, it means that the difference columns within the target template cluster are irrelevant to the patient's nursing needs. In this case, the common columns of all templates within the cluster can meet the patient's basic needs, and the presence or absence of difference columns has no substantial impact on the generation of the medical record. Therefore, an electronic nursing medical record template can be randomly selected from the target template cluster as the template for the new patient.

[0146] For example, if the target template cluster contains 3 templates, the difference column is "weight gain record", and the new patient is an adult patient, then any one of the 3 templates can be randomly selected, and its general columns can cover the patient's needs.

[0147] b. If the electronic nursing record of a new patient contains at least one discrepancy column in the test results, the electronic nursing record template with the highest number of discrepancy columns will be used as the electronic nursing record template for the new patient.

[0148] When the test results indicate that a new patient document contains at least one difference column, it means that these difference columns are directly related to the patient's care needs. The template with the most existing difference columns for these patients should be selected within the cluster to ensure that the template can fully carry the patient's care information.

[0149] For example, for each template within the target template cluster, the number of existing difference columns for new patients is counted. This number represents the intersection of the template's column list and the subset of difference columns for new patients. Further, the number of existing difference columns for new patients is compared across all templates, and the template with the largest number is selected as the electronic nursing record template for the new patient. If multiple templates have the same value for the number of existing difference columns for new patients, one can be randomly selected.

[0150] Based on the above technical solution, the embodiments of the present invention further narrow down the scope within the target template cluster by identifying differences, matching detection, and classifying selection, so as to ensure that the finally selected electronic nursing medical record template can not only match the nursing characteristics of new patients, but also cover their actual needs to the greatest extent, providing a precise framework for information filling of the medical record generation module 15.

[0151] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0152] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for generating electronic medical records for general nursing care, characterized in that, The method includes: The process involves acquiring historical electronic nursing documents and performing word segmentation; wherein, the historical electronic nursing documents are electronic nursing documents converted from historical paper nursing documents of patients. Based on the historical electronic nursing documents, an importance value and a core word segmentation array are determined for each type of electronic nursing document; wherein, the importance value is used to characterize the contribution weight of the corresponding type of electronic nursing document in constituting the electronic nursing medical record, and the core word segmentation array is used to characterize the key information features in the corresponding type of electronic nursing document; Electronic nursing record templates are clustered based on the importance value and core word segmentation array of each type of electronic nursing document to obtain multiple template clusters; wherein, each of the multiple template clusters contains electronic nursing record templates with similar structure and key features; Based on the new patient's paper nursing documents and the multiple template clusters, determine the new patient's electronic nursing record template from multiple electronic nursing record templates; The electronic nursing record of the new patient is determined based on the electronic nursing document and the electronic nursing record template of the new patient.

2. The method for generating electronic medical records for general nursing care according to claim 1, characterized in that, The acquisition of historical electronic nursing documents specifically includes: Obtain the paper-based nursing documents of the historical patients; wherein, the paper-based nursing documents include various types of documents; Convert the paper-based nursing documents into the historical electronic nursing documents; Each section of the electronic nursing document is segmented into words to obtain a segmented word array for each section.

3. The method for generating electronic medical records for general nursing care according to claim 1, characterized in that, The step of determining the importance value of each type of electronic nursing document based on the historical electronic nursing documents specifically includes: For each electronic nursing record template, obtain the set of historical patients corresponding to the electronic nursing record template; Obtain the number of times each type of electronic care document appears in the historical patient set, and determine the importance value of each type of electronic care document.

4. The method for generating electronic medical records for general nursing care according to claim 1, characterized in that, Based on the historical electronic nursing documents, the core word segmentation array for each type of electronic nursing document is determined, specifically including: For each type of electronic nursing document, the contribution value of each word in each section is calculated; wherein, the contribution value is used to characterize the frequency of occurrence of the word in the corresponding type of electronic nursing document; From the word segments of each column, select a preset number of word segments as core word segments in descending order of their contribution values. Based on the combination of core words formed by each column, determine the core word array for each type of electronic nursing document.

5. The method for generating electronic medical records for general nursing care according to claim 1, characterized in that, The process of clustering electronic nursing record templates based on the importance value and core word segmentation array of each type of electronic nursing document specifically includes: Based on the difference in importance values ​​of corresponding electronic nursing documents between two electronic nursing medical record templates and the core word segmentation array, the similarity value between each two electronic nursing medical record templates is determined. Based on the preset clustering algorithm and the similarity value between every two electronic nursing medical record templates, all electronic nursing medical record templates are clustered to obtain the multiple template clusters.

6. The method for generating electronic medical records for general nursing care according to claim 1, characterized in that, The step of determining the electronic nursing record template for the new patient from multiple electronic nursing record templates based on the new patient's paper nursing documents and the multiple template clusters specifically includes: The paper nursing documents of the new patient are obtained and converted and segmented to obtain the electronic nursing documents of the new patient. Determine the matching degree value between the electronic nursing documents of the new patient and each template cluster; The template cluster with the highest matching degree value is determined as the target template cluster, and the electronic nursing record template of the new patient is determined from the electronic nursing record templates contained in the target template cluster.

7. The method for generating electronic medical records for general nursing care according to claim 6, characterized in that, The step of determining the electronic nursing record template for the new patient from the electronic nursing record templates included in the target template cluster specifically includes: Identify the differences between the various electronic nursing record templates within the target template cluster; The electronic nursing document of the new patient is checked to see if it contains the aforementioned discrepancy section, and the results are obtained. Based on the test results, the electronic nursing record template for the new patient is determined.

8. The method for generating electronic medical records for general nursing care according to claim 7, characterized in that, The step of determining the electronic nursing record template for the new patient based on the test results specifically includes: If the detection results indicate that the new patient's electronic nursing record template does not contain the difference column, the new patient's electronic nursing record template is randomly selected from the target template cluster. If the electronic nursing document for the new patient contains at least one of the aforementioned discrepancy columns as indicated by the test results, the electronic nursing record template containing the highest number of discrepancy columns will be used as the electronic nursing record template for the new patient.

9. The method for generating electronic medical records for general nursing care according to claim 1, characterized in that, The step of determining the electronic nursing record based on the new patient's electronic nursing document and the new patient's electronic nursing record template specifically includes: Information is extracted from electronic nursing documents using natural language processing technology; The extracted information is mapped to the corresponding columns of the new patient's electronic nursing record template to generate the new patient's electronic nursing record.

10. An electronic medical record generation system for general practice nursing, characterized in that, The electronic medical record generation system for general nursing includes: a document processing module, a feature calculation module, a template clustering module, a template selection module, and a medical record generation module; The document processing module is used to acquire historical electronic nursing documents and perform word segmentation processing; wherein, the historical electronic nursing documents are electronic nursing documents converted from historical patients' paper nursing documents; The feature calculation module is used to determine the importance value and core word segmentation array of each type of electronic nursing document based on the historical electronic nursing documents; wherein, the importance value is used to characterize the contribution weight of the corresponding type of electronic nursing document in the formation of electronic nursing medical records, and the core word segmentation array is used to characterize the key information features in the corresponding type of electronic nursing document; The template clustering module is used to cluster electronic nursing medical record templates according to the importance value of each type of electronic nursing document and the core word segmentation array to obtain multiple template clusters; wherein, each of the multiple template clusters contains electronic nursing medical record templates with similar structures and key features; The template selection module is used to determine the electronic nursing record template for the new patient from multiple electronic nursing record templates based on the new patient's paper nursing documents and the multiple template clusters. The medical record generation module is used to determine the new patient's electronic nursing record based on the new patient's electronic nursing document and the new patient's electronic nursing medical record template.

Images