A diagnosis and treatment project information standardization processing method, storage medium and device

Abstract

The application relates to a diagnosis and treatment project information standardization processing method, a storage medium and equipment, and the method comprises the following steps: acquiring multi-source data, and extracting diagnosis and treatment project information in the multi-source data; performing data cleaning on the diagnosis and treatment project information to obtain a data structure table; adopting fusion of multiple matching algorithms, intelligently mapping the data structure table according to a standardized diagnosis and treatment project library, acquiring a mapping result between diagnosis and treatment projects in the data structure table and diagnosis and treatment project standard codes in the standardized diagnosis and treatment project library; and updating the mapping result based on artificial feedback information. Compared with the prior art, the application has the advantages of improving mapping efficiency and adaptability.

Description

Technical Field

[0001] This invention relates to the field of medical information processing technology, and in particular to a standardized processing method, storage medium, and device for diagnostic and treatment information. Background Technology

[0002] In the medical field, diagnostic and treatment procedures refer to the various measures taken by doctors to diagnose and treat diseases, such as examinations, tests, surgeries, and medications. The number and types of diagnostic and treatment procedures are numerous, and different hospitals and regions may have different naming and coding rules, which poses difficulties for the exchange and statistical analysis of medical information. Diagnostic and treatment procedure coding is a way for medical institutions to classify and identify the diagnostic and treatment services they provide; it is the foundation for medical quality management, medical cost accounting, and medical insurance payment. Diagnostic and treatment procedure standardization classification tools are tools for standardizing and classifying diagnostic and treatment procedures in medical practice. They can help medical institutions improve the quality and efficiency of diagnosis and treatment, control medical costs, and optimize the utilization of medical resources. They can also help insurance claims personnel improve the efficiency of claims processing and control claims risks. The application scenarios of diagnostic and treatment procedure standardization classification tools are mainly in insurance claims and medical and elderly care service scenarios, including the following aspects: 1) Disease Management in Medical Institutions: By selecting common, high-incidence, and resource-intensive diseases, standardized treatment packages are developed, thereby identifying, analyzing, and evaluating these treatment packages to achieve full-process management and quality control of the diseases.

[0003] 2) Medical Institution Expense Management: By classifying and analyzing the costs of standardized medical services, we can identify differences in cost composition and potential cost control opportunities, providing data support for medical insurance payment policies and hospital internal incentive mechanisms.

[0004] 3) Intelligent assistance for insurance claims: Utilizing artificial intelligence technology, the system standardizes the medical items in the claim details and performs functions such as intelligent early warning and intelligent analysis to assist claims personnel in making claims decisions and improve claims efficiency.

[0005] Inconsistencies, lack of standardization, and incompleteness in the coding of medical services across different regions, levels, and types of medical institutions have led to difficulties in medical information exchange and sharing, impacting the efficiency and quality of medical services. To address this issue, standardized medical service classification tools have emerged. These tools primarily resolve the following pain points: 1) Low efficiency in the collection and analysis of medical data. To assess medical expense claims and control, as well as the quality and effectiveness of treatment, a large amount of medical data needs to be collected and analyzed. This data contains crucial information such as treatment items, which are closely related to medical costs and treatment outcomes. However, due to the imperfections and lack of uniformity in hospital information systems, treatment items often refer to different coding standards and follow different charging rules. The data is scattered and non-standardized, making effective integration and comparison difficult. Furthermore, the non-standard and non-regulated formats of multi-source data also pose challenges to data analysis.

[0006] 2) The development and implementation of clinical pathways are challenging. A clinical pathway refers to establishing standardized treatment models and procedures for a specific target disease, standardizing the key diagnostic and treatment activities involved. However, the development of clinical pathways requires reference to multiple clinical guidelines and evidence-based medicine, and necessitates continuous updates and improvements. The implementation of clinical pathways also faces numerous challenges, such as non-standardized diagnosis and treatment, low homogeneity, inappropriate laboratory tests, and inappropriate drug use. Standardization of diagnostic and treatment items and even processes is required based on data standardization.

[0007] 3) There is a lack of effective means to improve the technical capabilities of medical network management. Currently, there is a lack of unified evaluation standards and indicators for medical network management, which makes it difficult to effectively quantify and objectify medical network management, and also makes it difficult to identify and solve existing problems and deficiencies.

[0008] Currently, the standardization of diagnostic and treatment items generally involves two steps. First, based on rule engines, natural language processing, machine learning, and other methods, the codes and names of diagnostic and treatment items are mapped to standard codes and names. Then, medical personnel manually remap any unsuccessful or incorrect codes and names. However, existing diagnostic and treatment item standardization classification tools still have some shortcomings or deficiencies, such as: 1) Mapping efficiency issues: Due to the complexity and diversity of diagnostic and treatment data, the involvement of natural language processing, and the inherent limitations of artificial intelligence methods, a certain error rate exists during the mapping process. This places higher demands on manual mapping: on the one hand, manual mapping of diagnostic and treatment items that fail to map successfully is required; on the other hand, the incorrectly mapped diagnostic and treatment items must be reviewed and revised.

[0009] 2) Data quality issues: Due to the complex and diverse sources of medical data, there are problems such as incompleteness, inaccuracy, inconsistency, and non-standardization, which affect the effectiveness and reliability of data analysis.

[0010] 3) Algorithm adaptability issues: As medical knowledge and clinical practice are constantly updated and changed, the mapping methods generated based on rules or algorithms may have problems such as lag, simplicity, and rigidity, affecting the timeliness and relevance of the data.

[0011] 4) Algorithm interpretability issues: Due to the inherent complexity and black-box nature of artificial intelligence technology, standardized classification tools designed based on statistical or machine learning techniques may have problems such as bias, error, and overfitting, affecting the validity and interpretability of the data. Summary of the Invention

[0012] The purpose of this invention is to overcome the shortcomings of the prior art by providing a method, storage medium, and device for standardizing the processing of diagnostic and treatment information to improve mapping efficiency and adaptability.

[0013] The objective of this invention can be achieved through the following technical solutions: A method for standardizing medical service information processing includes the following steps: Acquire multi-source data and extract diagnostic and treatment information from the multi-source data; The diagnostic and treatment information is cleaned to obtain a data structure table; By employing a fusion of multiple matching algorithms, the data structure table is intelligently mapped based on a standardized diagnostic and treatment item library to obtain the mapping results between the diagnostic and treatment items in the data structure table and the standard codes of the diagnostic and treatment items in the standardized diagnostic and treatment item library; The mapping results are updated based on human feedback.

[0014] Furthermore, the multi-source data includes various sources such as electronic medical records, medical order systems, and detailed expense lists.

[0015] Furthermore, the data cleaning of the aforementioned diagnostic and treatment information specifically includes: Locate the diagnosis and treatment item codes and content in the diagnosis and treatment item information, perform text processing, and obtain a data structure table. The text processing includes multiple methods such as formatting, deduplication, noise reduction, and completion.

[0016] Furthermore, the intelligent mapping includes: Based on a pre-established medical-specific lexicon, a professional medical word segmentation tool is used to segment the vocabulary of each diagnosis and treatment item in the data structure table and extract the corresponding key features. Using semantic similarity calculation, the extracted key features are matched with the corresponding features of each standard item in the standardized diagnosis and treatment item library. Based on the semantic similarity score, an initial candidate mapping list sorted in descending order of similarity is generated for each diagnosis and treatment item detail. For each candidate mapping in the initial candidate mapping list, perform context association verification, calculate the context consistency score, and exclude candidate mappings with a context consistency score lower than a preset value; The remaining candidate mappings are processed using predefined business rules and a historical mapping rule library to generate rule matching scores; The semantic similarity score, context consistency score, and rule matching score are weighted and summed to obtain a comprehensive score for each candidate mapping. The candidate mapping with the highest comprehensive score is selected as the output of the intelligent mapping. If the highest score is lower than the preset threshold, or if there are multiple candidate mapping scores that are close and all higher than the preset threshold, it is determined to be an ambiguous or polysemous case, and the mapping task is transferred to the manual matching queue.

[0017] Furthermore, the medical-specific thesaurus includes at least standard medical terms, commonly used clinical abbreviations, synonyms, near-synonyms, and local colloquialisms.

[0018] Furthermore, the medical-specific thesaurus is dynamically updated based on the human feedback information.

[0019] Furthermore, the key features include at least the project name, operating location, technical method, and unit of measurement.

[0020] Furthermore, the method also includes: Update the historical mapping rule library based on the mapping results.

[0021] The present invention also provides a computer-readable storage medium including one or more programs executable by one or more processors of an electronic device, said one or more programs including instructions for performing the diagnostic and treatment information standardization processing method described above.

[0022] The present invention also provides an electronic device, including one or more processors, a memory, and one or more programs stored in the memory, said one or more programs including instructions for performing the diagnostic and treatment information standardization processing method described above. Compared with the prior art, the present invention has the following beneficial effects: 1. This invention employs a mapping process that integrates multiple matching algorithms. This combines the advantages of different algorithms, improving matching accuracy through cross-validation or weighted judgment, reducing the possibility of mapping errors, and decreasing the workload of manual review. Simultaneously, based on the Pareto principle (80 / 20 rule), intelligent matching efficiently solves at least 80% of the standardization problems in medical procedures, improving the efficiency of hospital operational analysis and insurance claims calculation.

[0023] 2. In this invention, whenever the system completes a new mapping task, the matching logic, feature patterns, etc. generated during the mapping process will be automatically converted into reusable mapping rules and added to the historical mapping rule library. After the reviewers confirm, correct or supplement the mapping results, the manual feedback will be learned by the system and also be precipitated as new mapping rules. Through the rule precipitation method, new mapping tasks are continuously precipitated as mapping rules, ensuring the updating of the historical mapping rule library and adapting to new mapping requirements. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of the process of the present invention; Figure 2 This is a schematic diagram of the diagnosis and treatment information extracted from multi-source data in an embodiment of the present invention; Figure 3 This is a schematic diagram of approved diagnostic and treatment items in an embodiment of the present invention; Figure 4 This is a schematic diagram of the workflow of the diagnostic and treatment item coding standardization and classification tool in an embodiment of the present invention; Figure 5 This is a schematic diagram of the processing system in an embodiment of the present invention. Detailed Implementation

[0025] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. These embodiments are based on the technical solution of the present invention and provide detailed implementation methods and specific operating procedures. However, the scope of protection of the present invention is not limited to the following embodiments.

[0026] Example 1 This embodiment provides a method for standardizing the processing of medical service information, such as... Figure 1 As shown, it includes the following steps: S1. Obtain multi-source data and extract diagnostic and treatment information from the multi-source data, such as... Figure 2 As shown. The multi-source data includes electronic medical records, medical order systems, and detailed expense lists.

[0027] S2. Perform data cleaning on the diagnostic and treatment information to obtain a data structure table.

[0028] Specifically, data cleaning of medical treatment information includes: Locate the diagnosis and treatment item codes and content in the diagnosis and treatment item information, perform text processing, and obtain a data structure table. The text processing includes multiple methods such as formatting, deduplication, noise reduction, and completion.

[0029] Furthermore, formatting includes case conversion, full-width / half-width conversion, etc.

[0030] S3. By employing a fusion of multiple matching algorithms, the data structure table is intelligently mapped according to the standardized diagnosis and treatment item library to obtain the mapping results between the diagnosis and treatment items in the data structure table and the standard codes of the diagnosis and treatment items in the standardized diagnosis and treatment item library.

[0031] Specifically, intelligent mapping includes: Using a professional medical word segmentation tool, the details of each diagnosis and treatment item in the data structure table are segmented into words, and the corresponding key features are extracted; Based on a medical-specific thesaurus and the aforementioned key features, the details of diagnosis and treatment items are matched with the standard codes of diagnosis and treatment items in the standardized diagnosis and treatment item database.

[0032] Furthermore, based on text matching, natural language processing (NLP) techniques, and optimization rules, and leveraging a medically specific feature model and lexicon, detailed diagnostic and treatment data are mapped automatically and with high accuracy. First, a comprehensive and medically specific lexicon is established. This lexicon not only includes standard medical terminology but also integrates commonly used clinical abbreviations, synonyms, near-synonyms, and local idioms. It is continuously enriched through databases of actual diagnostic and treatment items used in various regions, ensuring comprehensive coverage of all types of diagnostic and treatment descriptions and providing a sufficient semantic foundation for subsequent text matching. This lexicon supports professional word segmentation and feature extraction, enhancing the ability to distinguish between ambiguities and polysemy. Second, medically-oriented NLP techniques are used for text parsing and feature extraction: professional medical word segmentation tools (such as jieba with a custom medical dictionary) are used, combined with the linguistic characteristics of the diagnostic and treatment descriptions, to segment words and extract key features, including structured information such as item names, operating sites, technical methods, and units of measurement. The feature extraction process emphasizes semantic understanding within a medical context to enhance the ability to parse complex expressions. Furthermore, based on the above lexicon and feature extraction results, an artificial intelligence mapping algorithm is used to match the details of diagnosis and treatment items with the standard codes of medical insurance diagnosis and treatment items. The intelligent algorithm, which is constructed by comprehensively using semantic similarity calculation, context association analysis and rule reasoning, achieves accurate mapping between non-standard expressions and standard codes, and supports the identification and handling of ambiguity or polysemy.

[0033] In this embodiment, the specific steps for achieving accurate mapping between non-standard expressions and standard codes by comprehensively utilizing intelligent algorithms such as semantic similarity calculation, context association analysis, and rule reasoning include: Step 1: Preprocessing and Feature Extraction Based on the established medical-specific lexicon, a medical word segmentation tool was used to segment the input diagnostic and treatment item details text and extract key features, including structured information such as item name, operation site, technical method, and unit of measurement.

[0034] Step 2: Candidate Set Generation Using semantic similarity calculations (such as cosine similarity, edit distance, or BERT-based vector coding), the extracted key features are quickly matched with the corresponding features of each standard item in the standardized diagnosis and treatment item library (medical insurance diagnosis and treatment item standard coding library) to generate semantic similarity scores, and an initial candidate mapping list sorted in descending order of similarity is generated for each non-standard expression.

[0035] In this embodiment, cosine similarity is used to calculate semantic similarity, and the calculation formula is as follows: This indicates the input text. This is a vector representation of the input text. This is a vector representation of a standard item in a standardized medical item library.

[0036] Cosine similarity is used as the semantic similarity score, and the Top-K candidate set is selected to form the initial candidate mapping list.

[0037] In other embodiments, when word granularity is insufficient, it can be upgraded to soft alignment, which updates the semantic similarity score based on a weighted sum of soft alignment score and cosine similarity.

[0038] Step 3: Contextual Analysis For each mapping pair in the candidate mapping list, contextual consistency verification is performed, such as checking whether features like the operation location, technical method, and unit of measurement are consistent with the applicable scope of the standard encoding. Using synonyms, abbreviations, and common collocations maintained in the thesaurus, the reasonable mapping direction of non-standard expressions in specific contexts is determined. A contextual consistency score is calculated, and obviously unreasonable candidates are eliminated.

[0039] Step 4: Rule-based reasoning Predefined business rules are applied to filter, score, or de-weight candidate mappings. Examples of rules include: if the unit of measurement does not match the unit required by the standard coding, the candidate is directly eliminated; if the same or highly similar non-standard representation mapping records exist in the historical mapping rule base, the mapping result is directly adopted. The result of rule reasoning generates a rule matching score.

[0040] Step 5: Comprehensive Scoring and Mapping Decision The semantic similarity score from step two, the context consistency score from step three, and the rule matching score from step four are weighted and summed to obtain a comprehensive score for each candidate mapping. The candidate with the highest comprehensive score is selected as the intelligent mapping output.

[0041] If the highest score is lower than the preset threshold, or if there are multiple candidate scores that are close to and all higher than the preset threshold, it is determined to be an ambiguous or polysemous case, and the mapping task is transferred to the manual matching queue.

[0042] Step Six: Results Output and Feedback Consolidation The final mapping results (including intelligent mapping output or correction results after manual review) are stored in the mapping result table of the data structure table and the standardized diagnosis and treatment item library. At the same time, manual feedback information (such as corrected mapping pairs, newly added rules, and supplementary terms) is stored in the mapping rule library and the thesaurus.

[0043] In this embodiment, the identification and handling of ambiguity or polysemy are specifically as follows: Situation assessment: 1) Low confidence judgment: If the highest comprehensive score calculated by intelligent mapping is lower than the preset threshold (0.7), it means that the algorithm is not confident in the correctness of the current match.

[0044] 2) High competitiveness judgment: If the comprehensive scores of multiple candidate mappings (two or more) are all higher than the threshold and the score difference is less than the set tolerance range (0.05), it means that the algorithm has identified multiple possible standard codes that are highly similar to the input expression and cannot be uniquely determined.

[0045] Handling strategy: 1) Automatic labeling and manual matching: All mapping tasks judged as ambiguous / multiple meanings will not directly output automatic results, but will be marked as "awaiting manual matching" and pushed to the manual matching queue.

[0046] 2) Providing auxiliary information: When transferring to human assistant, a list of candidate mappings and their scores are displayed, and key feature differences that cause ambiguity are highlighted.

[0047] 3) Feedback and retention of ambiguous results: After the manual reviewer completes the correction, the processing result of the ambiguous case (correct mapping to ambiguous features) will be recorded and used for automatic reasoning or rule updates in subsequent similar scenarios.

[0048] S4. Update the mapping result based on human feedback information.

[0049] In other embodiments, the method further includes: reviewing the data mapped by both intelligent and manual methods, and exporting the codes, names, and categories of the approved diagnostic and treatment items, such as... Figure 3 As shown.

[0050] The aforementioned intelligent matching algorithm, based on natural language processing, can accurately and automatically recommend the most suitable coding standards and classification methods according to different medical service item names. This improves the accuracy and consistency of coding classification and reduces the cost and error rate of manual coding. The method also features self-learning, self-adaptation, and manual adjustment steps. After matching, it continuously enriches the intelligent mapping rule table based on the results of manual review and mapping. This allows the rule base and matching algorithm to be continuously updated and optimized through intelligent and manual means according to the development and changes in the medical field, ensuring the timeliness and advancement of the coding standardization and classification tool and promoting the continuous accumulation of the tool's rule base. This method can improve the efficiency and quality of medical information management, unify and standardize the coding and classification of medical service items, facilitate data exchange and settlement between medical institutions and medical insurance departments, and promote the rational allocation and utilization of medical resources.

[0051] If the above methods are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0052] Example 2 This embodiment provides a standardized classification tool for medical item coding, used to implement the standardized processing method for medical item information as described in Embodiment 1.

[0053] In this embodiment, the deployment steps of the diagnostic and treatment item coding standardization classification tool are as follows: 1) Install a standardized classification tool for medical service item coding in the medical institution operation analysis system; 2) Import the diagnostic and treatment data that needs to be mapped into the diagnostic and treatment item standardization and classification tool, start the mapping task, complete the data cleaning, and wait for the intelligent mapping results to be generated; 3) Review and audit the mapping results in the standardized classification tool for diagnosis and treatment items, manually map the missing content in the mapping results, and modify or delete unreasonable or inaccurate mapping results; 4) Export the mapping results from the standardized classification tool for diagnosis and treatment items, and import the mapping results into the medical institution operation analysis system by selecting either interface or file method.

[0054] By following the steps above, the standardization process for classifying medical services can be completed. This tool can greatly simplify and improve the efficiency of standardization, avoid errors and omissions from manual operation, and ensure the quality and accuracy of standardization.

[0055] refer to Figure 4 The diagram illustrates the working process of the above-mentioned diagnostic and treatment item coding standardization and classification tool. Here, 0.1~0.5 represent data processing steps, S1~S6 represent data storage objects, and D1~D11 represent data streams. Specifically: 0.1 Medical Item Information Collection: Raw medical item information, including item name, code, price, and specifications, is obtained from various healthcare information systems and stored in the raw medical item information S1. The collected medical item information D1 is then output. This information may contain inconsistencies, incompleteness, or inaccuracies, requiring subsequent cleaning and matching. This step is the starting point of the data flow diagram and serves as the input to the entire system.

[0056] 0.2 Medical Service Information Cleaning: According to cleaning rule D2, the collected medical service information D1 is formatted, deduplicated, denoised, and completed to obtain cleaned medical service information D3, which is then stored in the cleaned medical service information S2. Cleaning rules can be tailored to specific needs, such as removing spaces, parentheses, and other special characters from names, and replacing commas with periods in prices. This step ensures data quality and improves data accuracy and completeness.

[0057] 0.3 Matching of Medical Item Information: Based on the standardized medical item database D4 and matching rules D5, the cleaned medical item information D3 is intelligently matched and manually matched to obtain intelligently matched medical item information D6 and manually matched medical item information D7, which are then stored in the intelligently matched medical item information S3 and manually matched medical item information S4, respectively. The standardized medical item database is the National Healthcare Security Administration's standard medical item information database. The matching rules are determined through full matching and merged matching to ensure matching accuracy. This step ensures data comparability and improves data uniformity and standardization.

[0058] 0.4 Medical Item Information Coding: According to coding rule D8, the matched medical item information D7 is coded to obtain coded medical item information D9, which is then stored in the coded medical item information S5. The coding rule is formulated according to the National Healthcare Security Administration's standard medical item information coding, and the coded medical item information can be easily exchanged and shared. This step ensures data retrieval and improves data simplicity and usability.

[0059] 0.5 Classification of Medical Service Item Information: Based on classification rule D10, the coded medical service item information D9 is classified to obtain classified medical service item information D11, which is then stored in the classified medical service item information S6. The classification rules are mainly formulated according to the hospital's cost analysis needs, and mainly include primary classifications such as surgery fees and examination fees, secondary classifications such as surgical fees for a certain system, and detailed tertiary classifications. The classified medical service item information can be easily imported into the medical institution operation analysis system for data analysis and statistics.

[0060] The diagnostic and treatment item coding standardization and classification tool in this embodiment can automatically identify, convert, compare, and standardize the diagnostic and treatment item codes of different medical institutions. It can help medical institutions achieve consistency and interoperability in diagnostic and treatment item classification, improve the accuracy and usability of medical data, and thus bring the following value to medical services: 1) Improve data standardization. By using standardized classification tools, the medical insurance coding standard is unified into a common language for medical insurance information exchange in the new era, thus achieving the standardization of medical insurance coding and promoting the implementation of medical insurance standard coding.

[0061] 2) Improve the level of medical quality management. Standardized classification tools can be used to effectively compare and evaluate the quality of medical services provided by different medical institutions, promote the formulation and supervision of medical quality standards, and improve medical safety and satisfaction.

[0062] 3) Optimize the medical expense accounting process. By using standardized classification tools, accurate accounting and analysis of medical expenses from different medical institutions can be achieved, avoiding cost errors and disputes caused by classification differences, and improving the rationality and transparency of medical expenses.

[0063] 4) Supports medical insurance payment services. Standardized classification tools enable rapid adaptation and execution of medical insurance payment rules for different medical institutions, simplifying the medical insurance settlement process, reducing medical insurance review costs, and improving medical insurance payment efficiency.

[0064] 5) Promote the exchange and sharing of medical information. Through standardized classification tools, seamless connection and transmission of diagnosis and treatment information from different medical institutions can be achieved, breaking down information silos and enabling cross-institutional sharing of patient information, thus providing patients with better continuous care.

[0065] Example 3 This embodiment provides a standardized processing system for medical service item information, such as... Figure 5 As shown, it includes a database server, an application server, and a client, among which, Database server: Stores a database of medical insurance information business coding standards, including medical service item classification and code standards, as well as the mapping relationship between local medical insurance catalog codes and national codes. The database server uses high-performance, highly reliable, and highly secure hardware and software systems to ensure data integrity, consistency, and availability.

[0066] Application server: Used to implement the standardized processing method for medical item information as described in Example 1, providing core functions including data import, data query, data review, data update, and data export. The application server adopts a distributed architecture, supports concurrent access by multiple users, and improves the system's response speed and processing capacity. The application server and the database server are connected via a high-speed network to achieve rapid data exchange and synchronization.

[0067] Client-side: Provides a user interface for standardized classification tools for diagnostic and treatment items, facilitating user operation and interaction. The client is a PC-based application that supports multiple operating systems and browsers. The client connects to the application server via the internet or a local area network to display and transmit data.

[0068] The implementation of the above application servers in business scenarios or business processes includes: Data collection: Collect structured medical information from various data sources such as electronic medical records, medical order systems, and detailed expense lists.

[0069] Data cleaning: Locate the codes and content of diagnosis and treatment items from multiple data sources, perform corresponding text processing such as case conversion, full-width / half-width conversion, and deletion of meaningless characters, and obtain a usable data structure table as the input field for mapping.

[0070] Data matching: Utilizing artificial intelligence technologies such as text matching and natural language processing, the system performs feature extraction and classification of medical insurance treatment item data based on the standard codes for these items, and then performs intelligent mapping. The intelligently mapped data is then manually mapped, and both the intelligently and manually mapped data undergo appropriate review.

[0071] Data Export: Export the mapping results, choosing either an interface or a file method, and import the mapping results into the medical institution operation analysis system.

[0072] The rest is the same as in Example 1.

[0073] The preferred embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection defined by the claims.

Claims

1. A method for standardizing the processing of medical service information, characterized in that, Includes the following steps: Acquire multi-source data and extract diagnostic and treatment information from the multi-source data; The diagnostic and treatment information is cleaned to obtain a data structure table; By employing a fusion of multiple matching algorithms, the data structure table is intelligently mapped based on a standardized diagnostic and treatment item library to obtain the mapping results between the diagnostic and treatment items in the data structure table and the standard codes of the diagnostic and treatment items in the standardized diagnostic and treatment item library; The mapping results are updated based on human feedback.

2. The method for standardizing medical item information processing according to claim 1, characterized in that, The multi-source data includes various sources such as electronic medical records, medical order systems, and detailed expense lists.

3. The method for standardizing medical item information processing according to claim 1, characterized in that, Data cleaning of the aforementioned diagnostic and treatment information specifically includes: Locate the diagnosis and treatment item codes and content in the diagnosis and treatment item information, perform text processing, and obtain a data structure table. The text processing includes multiple methods such as formatting, deduplication, noise reduction, and completion.

4. The method for standardizing medical item information processing according to claim 1, characterized in that, The intelligent mapping includes: Based on a pre-established medical-specific lexicon, a professional medical word segmentation tool is used to segment the vocabulary of each diagnosis and treatment item in the data structure table and extract the corresponding key features. Using semantic similarity calculation, the extracted key features are matched with the corresponding features of each standard item in the standardized diagnosis and treatment item library. Based on the semantic similarity score, an initial candidate mapping list sorted in descending order of similarity is generated for each diagnosis and treatment item detail. For each candidate mapping in the initial candidate mapping list, perform context association verification, calculate the context consistency score, and exclude candidate mappings with a context consistency score lower than a preset value; The remaining candidate mappings are processed using predefined business rules and a historical mapping rule library to generate rule matching scores; The semantic similarity score, context consistency score, and rule matching score are weighted and summed to obtain a comprehensive score for each candidate mapping. The candidate mapping with the highest comprehensive score is selected as the output of the intelligent mapping. If the highest score is lower than the preset threshold, or if there are multiple candidate mapping scores that are close and all higher than the preset threshold, it is determined to be an ambiguous or polysemous case, and the mapping task is transferred to the manual matching queue.

5. The method for standardizing medical item information processing according to claim 4, characterized in that, The medical-specific thesaurus includes at least standard medical terms, commonly used clinical abbreviations, synonyms, near-synonyms, and local colloquialisms.

6. The method for standardizing medical item information processing according to claim 4, characterized in that, The medical-specific terminology database is dynamically updated based on the human feedback information.

7. The method for standardizing medical item information processing according to claim 4, characterized in that, The key features include at least the project name, operating location, technical method, and unit of measurement.

8. The method for standardizing medical item information processing according to claim 4, characterized in that, The method also includes: Update the historical mapping rule library based on the mapping results.

9. A computer-readable storage medium, characterized in that, It includes one or more programs that are executed by one or more processors of an electronic device, the one or more programs including instructions for performing the standardized processing method for diagnostic and treatment information as described in any one of claims 1-8.

10. An electronic device, characterized in that, It includes one or more processors, a memory, and one or more programs stored in the memory, said one or more programs including instructions for performing the standardized processing method for diagnostic and treatment information as described in any one of claims 1-8.

Images