Automatic ICD coding optimization method and device based on medical ontology, electronic equipment and storage medium
By constructing an automatic optimization ICD coding method based on medical ontology, the problems of insufficient rare disease coding capabilities and high vocabulary mapping maintenance costs in existing technologies are solved, and highly accurate and reusable ICD coding is achieved with automatic optimization capabilities.
Patent Information
- Application Number
- CN202410378500.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-03-29
- Publication Date
- 2025-10-10
AI Technical Summary
Existing technologies are insufficient in processing ICD coding capabilities for uncommon or rare diseases, and are unable to resolve the large disparity in vocabulary similarity in medical texts. In addition, coding technologies based on vocabulary mapping have high maintenance costs and insufficient mapping portability.
An automatic optimization ICD coding method based on medical ontology is constructed. Through preprocessing, word segmentation, derivative word construction, medical terminology library matching and concept coding, the parent-child node relationship and attribute relationship in the medical ontology model are utilized to achieve effective mapping and automatic optimization of different versions of ICD codes.
It improves the coding hit rate of rare diseases, enhances the accuracy and reusability of ICD coding, reduces iteration costs, and has automatic optimization capabilities.
Smart Images
Figure CN120764482A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the field of medical data processing, and in particular to an automatic optimization ICD coding method and device based on medical ontology, an electronic device and a storage medium. BACKGROUND
[0002] International Classification of Disease (ICD) is an international unified disease classification method formulated by the World Health Organization (WHO) and is widely used in clinical research, medical outcome monitoring, health management and health resource allocation, etc. The accurate coding of ICD has a profound and extensive impact on the medical and health service system.
[0003] The prior art has the following disadvantages: the intelligent coding technology based on machine learning needs a large amount of annotated text data for pre-training and is insufficient in processing rare diseases or rare diseases; the coding technology based on simple short text similarity cannot solve the similarity problem of expression words with large gaps in medical text; the coding technology based on word table mapping has high maintenance cost and insufficient mapping portability. SUMMARY
[0004] In view of the above technical problems in the prior art, the present application provides an automatic optimization ICD coding method and device based on medical ontology, an electronic device and a storage medium to improve the above problems.
[0005] To achieve the above purpose, the embodiments of the present application provide the following technical solutions:
[0006] In one aspect, the present application provides an automatic optimization ICD coding method based on medical ontology, which comprises:
[0007] Obtaining medical data;
[0008] Preprocessing the medical data to obtain preprocessed medical data;
[0009] Matching the preprocessed medical data with a medical terminology library to obtain the concept coding of the preprocessed medical data;
[0010] According to the mapping relationship between the concept coding and the ICD coding, the ICD coding result of the preprocessed medical data is obtained.
[0011] Further, the preprocessing of the medical data to obtain the preprocessed medical data comprises:
[0012] Standardizing the medical data to obtain standardized medical data.
[0013] Constructing a BERT+BiLSTM+CRF training word segmentation model, and using the model to segment the standardized medical data to obtain the segmentation of the standardized medical data;
[0014] Obtaining derivatives of the standardized medical data according to a medical field terminology set;
[0015] The standardized medical data and the derivatives of the standardized medical data are merged to obtain preprocessed medical data.
[0016] Furthermore, matching the preprocessed medical data with a medical terminology database to obtain a concept code for the preprocessed medical data includes:
[0017] Construct a medical ontology model according to the OWL 2 EL standard;
[0018] Based on the medical ontology model, a medical field terminology set is constructed, including terms, concepts, and relationships; the medical field includes diseases, symptoms, signs, anatomy, surgery, biology, drugs, medical devices, testing and examination, imaging, nursing, genes, and gene mutations;
[0019] Based on the medical field terminology set, a term extension set is constructed, including an extended set of colloquial disease and symptom terms, an extended set of drug English terms, and an extended set of disease, symptom and sign English terms;
[0020] Construct a medical terminology database based on terms and term extension sets;
[0021] matching the pre-processed medical data with a medical terminology database to obtain the medical term with the highest matching degree;
[0022] According to the mapping relationship between medical terms and the medical field term set, the concept coding of the preprocessed medical data is obtained.
[0023] Furthermore, the concept code refers to a unique identifier that represents a concept in a set of medical terminology.
[0024] Furthermore, the ICD code includes any one of the following coding systems or a combination thereof:
[0025] ICD-11, ICD-10, ICD-9, ICD-9-CM-3, ICD-O, and custom ICD coding systems.
[0026] In a second aspect, the present invention provides an automatic ICD coding optimization device based on medical ontology, comprising:
[0027] An acquisition module, used for acquiring medical data to be encoded;
[0028] a preprocessing module, configured to preprocess the medical data to obtain preprocessed medical data;
[0029] a matching module, configured to match the preprocessed medical data with an ontology-based medical term or a medical term expansion set to obtain concept coding;
[0030] a mapping module, configured to obtain ICD coding according to a mapping relationship between the concept coding and the ICD coding.
[0031] In a third aspect, the present application provides a computer device, comprising:
[0032] a memory, configured to store a computer program;
[0033] a processor, configured to execute the computer program to implement the steps of the method according to any one of claims 1 to 5.
[0034] In a fourth aspect, the present application provides a computer readable storage medium, wherein the computer readable storage medium has a computer program, and the computer program is executed by a processor to implement the steps of the method according to any one of claims 1 to 5.
[0035] Compared with the prior art, the present application has the following beneficial effects:
[0036] A large-scale medical term library is formed by a medical field term set constructed based on a medical ontology model, the problem of recognizing medical synonyms with large differences in text expression is solved, and the accuracy of ICD coding is improved.
[0037] The parent-child node relationship and attribute relationship between medical terms in the medical ontology model are used to construct a derivative word and a term expansion set of medical data, the problem that a traditional segmentation network and an expansion term can only cover a small number of attribute relationships such as symptoms, parts, and causes is solved, the derivative ability of medical data is improved, and the coding hit rate of rare diseases is improved.
[0038] The mapping of concept coding and ICD coding realizes effective mapping of different versions of ICD coding and improves reusability.
[0039] Through the update and maintenance of the medical field term set, the ICD coding accuracy is continuously improved, the automatic optimization capability is achieved, and the iteration cost is reduced.
[0040] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art from the description, or can be learned by practice of the present application according to the embodiments. The objects and other advantages of the present application will be realized and achieved by the structure particularly pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.
[0042] Figure 1 is a flow chart of an embodiment of a method according to the present invention;
[0043] Figure 2 FIG. 1 is a block diagram of an apparatus according to an embodiment of the present invention. DETAILED DESCRIPTION
[0044] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.
[0045] The present invention will be further described below with reference to the accompanying drawings and embodiments.
[0046] Figure 1 As shown in FIG, the automatic optimization ICD coding method based on medical ontology includes the following steps:
[0047] Step S1: Acquire medical data.
[0048] Specifically, the medical data may be a group of keywords or a piece of text, but is not limited thereto.
[0049] Step S2: preprocessing the medical data to obtain preprocessed medical data.
[0050] Step S2 further includes step S21, step S22, step S23 and step S24:
[0051] Step S21: performing standardization processing on the medical data to obtain standardized medical data.
[0052] Specifically, the standardization process includes data cleaning of medical data, such as removing punctuation marks and non-alphanumeric characters.
[0053] Step S22: A BERT+BiLSTM+CRF training segmentation model is constructed, and the model is used to segment the standardized medical data, to obtain the segmentation of the standardized medical data.
[0054] Step S23: According to the medical field term set, the derived words of the standardized medical data are obtained.
[0055] Step S24: The standardized medical data and the derived words of the standardized medical data are merged to obtain preprocessed medical data.
[0056] Step S3: The preprocessed medical data obtained in step S2 is matched with the medical term library to obtain the concept coding of the preprocessed medical data.
[0057] Step S3 further includes steps S31, S32, S33, S34, S35 and S36:
[0058] Step S31: According to the OWL 2 EL standard, a medical ontology model is constructed.
[0059] Specifically, ontology is a concept model describing concepts and relationships between concepts, and the definition of a concept is described through the relationship between concepts; it mainly faces a specific field and is used to describe the concept model of a specific field. It has five basic modeling elements: classes, relations, functions, axioms and instances. Classes include the name of a concept, the set of relationships between other concepts, and the description of the concept in natural language; relations are the interactions between concepts in a field, such as child node relationships, and the semantic relationship corresponds to a set of object tuples; functions are a special kind of relationship, such as parent node relationship or mother node relationship; axioms represent tautological assertions, such as the concept B belonging to the scope of concept A; instances represent elements, and semantically, instances represent objects such as disease names or term names.
[0060] Step S32: According to the medical ontology model, a medical field term set is constructed, including terms, concepts and relationships; the medical field includes diseases, symptoms, signs, anatomy, surgery, biology, drugs, medical devices, laboratory tests, imaging, nursing, genes and gene mutations.
[0061] Specifically, the medical field term set is a set of medical terms that meet the clinical applicability. There are three types of core components: concepts with unique meanings; terms that describe concepts in line with clinical usage habits; and relationships that link each concept to other related concepts.
[0062] A term is a linguistic reference to a concept. Multiple terms can exist under one concept at the same time, and each term has a unique digital identifier.
[0063] Step S33: Based on the medical field terminology set, a term extension set is constructed, including an extended set of colloquial disease and symptom terms, an extended set of drug English terms, and an extended set of disease, symptom and sign English terms.
[0064] Specifically, colloquial diseases and symptoms are words or phrases with a colloquial nature, distinct from medical terminology. By establishing relationships between colloquial disease and symptom terms and concepts in the medical terminology, the connection between colloquial terms and professional terms can be achieved. The extended set of English terms for drugs establishes relationships between the English terms for drugs and concepts in the medical terminology, thereby achieving English-Chinese correspondence for drugs. The extended set of English terms for diseases, symptoms, and signs establishes relationships between the English terms for diseases, symptoms, and signs and concepts in the medical terminology, thereby achieving English-Chinese correspondence for diseases, symptoms, and signs.
[0065] Step S34: Construct a medical terminology database based on the terms and term extension sets.
[0066] Step S35: matching the pre-processed medical data with a medical terminology database to obtain medical terms with the highest matching degree.
[0067] Specifically, the semantic similarity of multiple matching results is calculated and ranked.
[0068] Step S36: Obtain the concept coding of the pre-processed medical data according to the mapping relationship between the medical terms and the medical field terminology set.
[0069] Step S4: Obtaining an ICD coding result for the preprocessed medical data based on a mapping relationship between the concept code and the ICD code. The ICD code includes any one of the following coding systems or a combination thereof: ICD-11, ICD-10, ICD-9, ICD-9-CM-3, ICD-O, and a custom ICD coding system.
[0070] According to the automatic optimization ICD coding method based on medical ontology according to the embodiment of the present invention, a large-scale medical terminology library is formed by a medical field terminology set constructed based on the medical ontology model, which solves the problem of identifying medical synonyms with large differences in text expression and improves the accuracy of ICD coding; the parent-child node relationship and attribute relationship between medical terms contained in the medical ontology model are used to construct derivative words and term extension sets of medical data, which solves the problem that traditional word segmentation networks and extended terms can only cover a small number of attribute relationships such as symptoms, parts, and causes, improves the derivation ability of medical data, and improves the coding hit rate of rare diseases; the mapping of concept codes and ICD codes realizes the effective mapping of different versions of ICD codes and improves reusability; through the update and maintenance of the medical field terminology set, the accuracy of ICD coding is continuously improved, with automatic optimization capabilities and reduced iteration costs.
[0071] like Figure 2 As shown, the automatic optimization ICD coding device based on medical ontology includes:
[0072] (1) Acquisition module, used to acquire medical data to be coded;
[0073] (2) a preprocessing module, configured to preprocess the medical data to obtain preprocessed medical data;
[0074] (3) a matching module for matching the preprocessed medical data with ontology-based medical terms or medical term extensions to obtain concept codes;
[0075] (4) Mapping module, used to obtain ICD codes based on the mapping relationship between concept codes and ICD codes.
[0076] In another embodiment, the present invention provides a computer device comprising a memory and a processor, wherein the memory is used to store a computer program; and the processor is used to implement the combined Figure 1 The steps of the method embodiments shown and described or other corresponding method embodiments are not repeated here.
[0077] In another embodiment, the present invention provides a computer readable storage medium having a computer program thereon, which when executed by a processor implements Figure 1 The steps of the method embodiments shown and described or other corresponding method embodiments are not repeated here.
[0078] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.
Claims
1. A method for automatically optimizing ICD coding based on medical ontology, characterized in that: include: Access to medical data; preprocessing the medical data to obtain preprocessed medical data; Matching the preprocessed medical data with a medical terminology database to obtain a concept code for the preprocessed medical data; According to the mapping relationship between the concept code and the ICD code, the ICD coding result of the preprocessed medical data is obtained.
2. The method for automatically optimizing ICD coding based on medical ontology according to claim 1, characterized in that: The preprocessing of the medical data to obtain the preprocessed medical data includes: performing standardization processing on the medical data to obtain standardized medical data; Constructing a BERT+BiLSTM+CRF training word segmentation model, and using the model to segment the standardized medical data to obtain the segmentation of the standardized medical data; Obtaining derivatives of the standardized medical data according to a medical field terminology set; The standardized medical data and the derivatives of the standardized medical data are merged to obtain preprocessed medical data.
3. The automatic optimization ICD coding method based on medical ontology according to claim 1 is characterized in that: The matching of the pre-processed medical data with a medical terminology database to obtain a concept code for the pre-processed medical data includes: Construct a medical ontology model according to the OWL 2 EL standard; Based on the medical ontology model, a medical field terminology set is constructed, including terms, concepts, and relationships; the medical field includes diseases, symptoms, signs, anatomy, surgery, biology, drugs, medical devices, testing and examination, imaging, nursing, genes, and gene mutations; Based on the medical field terminology set, a term extension set is constructed, including an extended set of colloquial disease and symptom terms, an extended set of drug English terms, and an extended set of disease, symptom and sign English terms; Construct a medical terminology database based on terms and term extension sets; matching the pre-processed medical data with a medical terminology database to obtain the medical term with the highest matching degree; According to the mapping relationship between medical terms and the medical field term set, the concept coding of the preprocessed medical data is obtained.
4. The method for automatically optimizing ICD coding based on medical ontology according to claim 1 and claim 3, characterized in that: The concept code refers to a unique identifier that represents a concept in a set of medical terminology.
5. The method for automatically optimizing ICD coding based on medical ontology according to claim 1, characterized in that: The ICD code includes any one of the following coding systems or their combination: ICD-11, ICD-10, ICD-9, ICD-9-CM-3, ICD-O, and custom ICD coding systems.
6. A device for automatically optimizing ICD coding based on medical ontology, characterized in that: include: An acquisition module, used for acquiring medical data to be encoded; A preprocessing module, configured to preprocess the medical data to obtain preprocessed medical data; a matching module, configured to match the preprocessed medical data with ontology-based medical terms or medical term extensions to obtain concept codes; The mapping module is used to obtain the ICD code according to the mapping relationship between the concept code and the ICD code.
7. A computer device, characterized in that: include: Memory for storing computer programs; A processor, configured to implement the steps of the method according to any one of claims 1 to 5 when executing the computer program.
8. A computer-readable storage medium, characterized in that The computer-readable storage medium carries a computer program, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.