Traditional Chinese medicine precision diagnosis and treatment model construction method based on real clinical data

Abstract

The invention provides a traditional Chinese medicine precision diagnosis and treatment model construction method based on real clinical data, and the method achieves the construction of a five-level precision diagnosis and treatment model of traditional Chinese medicine clinical specialist diseases' disease-stage-syndrome-syndrome-syndrome grouping-miscellaneous diseases' through a big data knowledge engineering technology. According to the method, massive traditional Chinese medicine clinical data is deeply treated, including data cleaning, standardization and knowledge-based processing, so that a comprehensive and accurate traditional Chinese medicine clinical case knowledge base is constructed; according to the method, diseases are further staging, syndromes are further subdivided into multiple groups on the basis of disease staging, and multi-dimensional data such as dialectical elements of each group and miscellaneous disease syndromes are analyzed, so that more refined diagnosis and treatment decision support is provided for traditional Chinese medicine clinic. The method can provide a certain basis for objective research of traditional Chinese medicine diagnosis and treatment, improves the implicit knowledge mining ability of traditional Chinese medicine clinical big data knowledge engineering, and has important clinical application value and wide market prospects.

Description

Technical Field

[0001] The present invention relates to the technical field of traditional medicine, and particularly to a method for constructing a precise diagnosis and treatment model of traditional Chinese medicine based on real clinical data. Background Art

[0002] The process of traditional Chinese medicine clinical practice is often divided into cognition, judgment, decision-making, and verification. Its model can generally be divided into three steps: "collection of four diagnostic information → syndrome differentiation and treatment → prescription (or other treatment methods)". In the early models, syndrome differentiation and treatment is a processing process of the doctor on the matching relationship between the four diagnostic information and the prescription drugs, and it is divided into two processes: syndrome differentiation and treatment. In recent years, natural language processing (NLP) based on medical records has gradually become a new research direction in the field of diagnosis and treatment model research. The research on diagnosis and treatment models is divided into five categories: finding intervention and diagnostic factors, establishing a diagnosis and treatment model without external verification, establishing a diagnosis and treatment model with external verification, model verification, and model impact research.

[0003] However, the diagnostic and treatment models constructed in the prior art have the following problems: 1. Currently, the research on diagnostic and treatment models based on medical records generally has problems such as fixed data forms, single data sources, extremely insufficient overall data volume and data dimensions; 2. Currently, the correlation calculation method of traditional Chinese medicine clinical diagnostic and treatment models strongly depends on using artificial methods to label features and weights, but the understanding levels and traditional Chinese medicine clinical thinking of each annotator are different, and the annotation results are different, which greatly affects the quality of data governance; 3. In the part of diagnostic basis, the currently studied diagnostic and treatment models basically only incorporate data from one type of guideline or textbook. Therefore, the trained models can only recognize the diseases, syndromes, treatments, and prescriptions in this version of the basis, with a small coverage area and low accuracy. The difference in the final discrimination results of the models mainly depends on the coverage area of the diagnostic basis; 4. The analytical research under the guidance of the reductionist thinking mode with the three-level diagnostic and treatment idea of disease, syndrome, and symptom as the core objectifies the fuzzy relationship of traditional Chinese medicine diagnosis and treatment, but there are also problems of insufficient meticulousness. Currently, many clinical specialties and diseases are more refined in clinical thinking. Detached from real traditional Chinese medicine clinical practice data, there will inevitably be one-sidedness, and even misleading, in the research results in explaining theoretical problems. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the purpose of the present invention is to provide a method for constructing a precise diagnosis and treatment model of traditional Chinese medicine based on real clinical data, so as to solve the problems raised in the above background art. The present invention first creates a five-layer hierarchical precise diagnosis and treatment model with a hierarchical structure of "disease differentiation → staging → syndrome differentiation → syndrome grouping → identification of concurrent and complex diseases, syndromes, and symptoms", in order to be more in line with the actual situation of traditional Chinese medicine clinical practice.

[0005] To achieve the above object, the present invention is implemented through the following technical solutions: A method for constructing a precise diagnosis and treatment model of traditional Chinese medicine based on real clinical data. The construction process of this method is divided into two parts: a diagnosis model and an intervention model. First, determine the specialized diseases and collect the relevant diagnosis and intervention criteria for the corresponding specialized diseases, construct a clinical diagnosis and treatment standard library, and form a standardized corpus, a syndrome entry pool, and an entry pool for each syndrome differentiation element of this specialized disease; classify, integrate, and normalize the collected clinical data of specialized diseases with the standardized corpus, and after knowledge encoding, form a clinical case knowledge base for specialized diseases. On this basis, construct a precise diagnosis model and an intervention model for this specialized disease.

[0006] Further, during the construction of the diagnosis model, it is necessary to determine the staging of each disease and the grouped diagnosis information of each syndrome.

[0007] Further, the construction of the intervention model is based on precise diagnosis to determine the core prescriptions and medication additions and subtractions for each syndrome group.

[0008] Further, it also includes the following steps of the standardized operation process for constructing the diagnosis and treatment model:

[0009] Step 1: Disease identification: After completing the collection of real clinical data and relevant standards for this specialized disease, classify and code the disease name according to the national standard "Classification and Codes of Traditional Chinese Medicine Syndromes" (GB / T 15657 - 2021).

[0010] Step 2: Disease staging: After determining the disease name, customize the disease staging principles and methods according to the research topic design and clinical actual situation of the disease being studied.

[0011] Step 3: Syndrome differentiation: According to the requirements and methods of traditional Chinese medicine clinical syndrome differentiation diagnosis, through comprehensive analysis of syndrome differentiation elements such as clinical symptoms, tongue manifestations, pulse conditions, and objective examination results, identify the syndrome to which the research belongs.

[0012] Step 4: Syndrome grouping: Obtain the core prescription information from the clinical data. If there is no core prescription information, determine it through cluster analysis of medication data, and at the same time, conduct syndrome grouping based on the usage of the core prescription.

[0013] Step 5: Determine syndrome differentiation elements: After disease staging and syndrome grouping, determine the syndrome differentiation elements of each syndrome group and establish a syndrome differentiation element database for them according to syndrome grouping.

[0014] Step 6: Identify concurrent diseases, syndromes, and symptoms: Based on the syndrome differentiation element database of each syndrome group, identify and extract the information of concurrent diseases, syndromes, and symptoms of each syndrome group, and establish a database for concurrent diseases, syndromes, and symptoms of this syndrome group according to three categories: concurrent diseases, concurrent syndromes, and concurrent symptoms.

[0015] Step 7: Determine the core prescription medications;

[0016] Step 8. Determine the additional and subtracted medications for complicated diseases, syndromes, and symptoms.

[0017] Further, in Step 1, if the disease name in the original data conforms to the coding standard of Classification and Codes of Traditional Chinese Medicine Diseases and Syndromes, it is coded based on this as the disease name.

[0018] Further, if the disease in the original data is expressed in a way that does not conform to the coding standard of Classification and Codes of Traditional Chinese Medicine Diseases and Syndromes, it is reclassified and coded according to clinical practice and expert consultation; if Western medicine disease names are used, the disease names should be determined and coded according to the disease classification of ICD.

[0019] Further, in Step 3, the syndrome classification and codes of the research disease types are determined according to the national standard Classification and Codes of Traditional Chinese Medicine Diseases and Syndromes.

[0020] Further, in Step 4, if there is no core prescription information in the clinical data, its medication data is used for cluster analysis to determine the core prescription for each disease syndrome, and the syndrome is grouped according to its different core prescriptions, and a syndrome grouping information database for this disease syndrome is established.

[0021] Further, Step 7 is used to determine the drug names and dosage information of each core prescription, and a core prescription medication information database for this disease syndrome is established according to syndrome grouping.

[0022] Further, in Step 8, the core prescription medications in the clinical prescription data are deleted to obtain the medication information corresponding to complicated diseases, syndromes, and symptoms, and then according to the functional indications of the drugs, the drugs corresponding to complicated diseases, syndromes, and symptoms are determined respectively.

[0023] Advantages of the present invention:

[0024] 1. Based on the traditional three-level diagnosis and treatment centered on "disease differentiation → syndrome differentiation → analysis of complicated syndromes", this method for constructing a precise traditional Chinese medicine diagnosis and treatment model based on real clinical data first creates a five-layer precise diagnosis and treatment model with a hierarchical structure of "disease differentiation → staging → syndrome differentiation → syndrome grouping → identification of complicated disease syndromes and symptoms", solving the problem that the current traditional diagnosis and treatment model does not fit well with clinical practice; in the application of diagnosis and treatment standards in the diagnosis and treatment model, the syndrome differentiation elements in the diagnosis and treatment standards of each clinical specialty and disease are applied to the establishment of each syndrome differentiation element in the form of an integral scale.

[0025] 2. This method of constructing a TCM precision diagnosis and treatment model based on real clinical data further stages the disease, subdivides the syndrome into syndrome classification, and retains multi-dimensional data such as syndromes of concurrent diseases, thereby providing more refined diagnosis and treatment decision support for TCM clinics. Through this method, the personalization and accuracy of TCM diagnosis and treatment can be significantly improved, treatment plans can be optimized, and clinical treatment effects can be improved. It has important clinical application value and broad market prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 A standardized operation flow chart for constructing the TCM clinical precision diagnosis and treatment model of the present invention;

[0027] Figure 2 This is the TCM precision diagnosis and treatment model of the present invention. DETAILED DESCRIPTION

[0028] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0029] See also Figures 1 to 2 , the present invention provides the following technical solutions: a method for constructing a precision diagnosis and treatment model of traditional Chinese medicine based on real clinical data, and its construction procedure is divided into two parts: a diagnosis model and an intervention model, and the two parts are distinguished from each other and supportive of each other. The construction of the precision diagnosis and treatment model of traditional Chinese medicine clinical practice first determines the specialty and disease, collects the relevant diagnosis and intervention standards for this specialty and disease, and constructs a clinical diagnosis and treatment standard library to form a standard corpus, a syndrome entry pool, and a syndrome differentiation element entry pool for this specialty and disease. The collected clinical data of specialties and diseases are classified, integrated, and normalized with the standard corpus, and a clinical case knowledge base of specialties and diseases is formed after knowledge encoding. On the basis of descriptive statistical analysis, a precision diagnosis model and intervention model for this specialty and disease are constructed. In the process of constructing the diagnosis model, it is necessary to determine the grouping diagnosis information of each disease stage and each syndrome. The construction of the intervention model is based on precise diagnosis to determine the core prescription and medication addition and subtraction for each syndrome grouping. The standardized operating procedure (SOP) for the construction of the precision diagnosis and treatment model of traditional Chinese medicine clinical practice is as follows: Figure 1 shown.

[0030] Through big data knowledge engineering technology, the present invention realizes the construction of a five-level precise diagnosis and treatment model of "disease-stage-syndrome-syndrome grouping-complicated syndrome and symptom" for specialized diseases in traditional Chinese medicine clinical practice. The core of this method lies in the in-depth governance of a large amount of traditional Chinese medicine clinical data, including data cleaning, standardization, and knowledge processing, to construct a comprehensive and accurate traditional Chinese medicine clinical case knowledge base; the innovation lies in further staging the disease, subdividing the syndrome into syndrome classifications, and retaining multi-dimensional data such as complicated syndrome and symptom, so as to provide more refined diagnosis and treatment decision-making support for traditional Chinese medicine clinical practice. Through this method, the personalization and accuracy of traditional Chinese medicine diagnosis and treatment can be significantly improved, the treatment plan can be optimized, and the clinical treatment effect can be enhanced, which has important clinical application value and broad market prospects.

[0031] This embodiment also provides a detailed discussion of the steps in each stage according to the levels shown in the standardized operation flowchart, which is as follows:

[0032] 1. Disease differentiation

[0033] After completing the collection of real clinical data and related standards for the specialized disease, determine the classification and code of the traditional Chinese medicine disease name according to the national standard "GB / T15657-2021 Classification and Codes of Traditional Chinese Medicine Diseases and Syndromes".

[0034] If the disease name in the original data conforms to the coding specification of "Classification and Codes of Traditional Chinese Medicine Diseases and Syndromes", code it according to the disease name; if the disease in the original data is expressed in a way that does not conform to the coding specification of "Classification and Codes of Traditional Chinese Medicine Diseases and Syndromes", re-classify and code it according to clinical practice and expert consultation. If a Western medicine disease name is adopted, determine the disease name according to the disease classification of ICD and code it.

[0035] 2. Disease staging

[0036] After determining the disease name, according to the research topic design and clinical actual situation of the disease studied, the disease staging principle and method can be customized. For example, the disease can be divided into acute stage, remission stage, recovery stage, etc., and then a disease staging information database is formed, which contains the judgment criteria and basis for disease staging of clinical data; or the disease can not be staged.

[0037] 3. Syndrome differentiation

[0038] According to the requirements and methods of traditional Chinese medicine (TCM) clinical syndrome differentiation diagnosis, through comprehensive analysis of syndrome differentiation elements such as clinical symptoms, tongue manifestations, pulse conditions, and objective examination results, identify the syndrome to which the research belongs, and determine the syndrome classification and code of the research disease type according to the national standard "Classification and Codes of TCM Diseases and Syndromes". When the TCM syndrome names in the original cases are inconsistent with the standardized TCM syndrome names, they should be matched according to the "Optional Words for TCM Syndrome Names" in the standard. If there are syndrome names that cannot be matched, the experts should standardize the syndrome names according to the "Classification and Codes of TCM Diseases and Syndromes" based on the clinical actual situation. Finally, form the "Clinical XX Specialty and Disease Syndrome Corpus Coding Database" with the TCM syndrome names and standardized syndromes in all original cases.

[0039] 4. Syndrome grouping

[0040] After determining the TCM disease classification and code of the research disease type, directly obtain the core prescription information from the clinical data. If there is no core prescription information in the clinical data, use its medication data for cluster analysis to determine the core prescription for each disease syndrome. Divide the syndrome into groups according to its different core prescriptions, and establish a syndrome grouping information database for this disease syndrome, such as databases for syndrome grouping 1, syndrome grouping 2... syndrome grouping N, etc.

[0041] 5. Determine syndrome differentiation elements

[0042] Symptom data is standardized and coded according to "Classification and Codes of Basic Clinical Symptom Information in TCM (T / CIATCM 020—2019)", "Classification and Codes of TCM Tongue Diagnosis Information (T / CIATCM 010—2019)", and "Classification and Codes of TCM Pulse Diagnosis Information (T / CIATCM 011—2019)", referring to Table 2 "Classification Name and Code Table of Key Symptoms" in them. If there are symptom names that cannot be matched, the experts should standardize the symptom names according to the "Classification and Codes of TCM Diseases and Syndromes" based on the clinical actual situation. After disease staging and syndrome grouping, determine the syndrome differentiation elements of each syndrome group, and establish a syndrome differentiation element database for it according to the syndrome grouping, such as syndrome differentiation element databases for syndrome grouping 1, syndrome grouping 2... syndrome grouping N.

[0043] 6. Identify concurrent diseases, syndromes, and symptoms

[0044] Based on the syndrome differentiation element database of each syndrome group, identify and extract the information of concurrent diseases, syndromes, and symptoms of each syndrome group, and establish a database of concurrent diseases, syndromes, and symptoms for this syndrome group according to the three categories of concurrent diseases, concurrent syndromes, and concurrent symptoms.

[0045] 7. Determine the medications of the core prescription

[0046] Determine the drug names and dosage information of each core prescription, and establish a core prescription medication information database for this disease syndrome according to the syndrome grouping.

[0047] 8. Determine the additional and subtracted medications for concurrent complex diseases, syndromes, and symptoms

[0048] Delete the core prescription medications in the clinical prescription data to obtain the medication information corresponding to concurrent complex diseases, syndromes, and symptoms, and then determine the medications corresponding to each of the concurrent complex diseases, syndromes, and symptoms according to the specifications of the functions and indications of the medications.

[0049] Use the knowledge association analysis algorithm to determine the association relationships between the concurrent complex diseases, syndromes, and symptoms and the medications in the data, and finally determine the additional and subtracted medication information corresponding to each of the concurrent complex diseases, syndromes, and symptoms of this syndrome group.

[0050] The above has shown and described the basic principles, main features, and advantages of the present invention. For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and without departing from the spirit or basic features of the present invention, the present invention can be implemented in other specific forms.

[0051] In addition, it should be understood that although this specification is described according to the embodiments, not every embodiment only contains an independent technical solution. This narrative way of the specification is only for clarity. Those skilled in the art should regard the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A method for constructing a precise traditional Chinese medicine diagnosis and treatment model based on real clinical data, characterized in that: The construction procedures of this method are divided into two parts: a diagnostic model and an intervention model. First, specific specialties and diseases are determined, relevant diagnostic and intervention criteria for the corresponding specialties and diseases are collected, a clinical diagnosis and treatment standard library is constructed, and a standardized corpus, a syndrome item pool, and an item pool for syndrome differentiation elements of each syndrome for this specialty and disease are formed. The clinical data of the specialty and disease collected are classified, integrated, and normalized with the standardized corpus, and after knowledge encoding, a clinical case knowledge base for the specialty and disease is formed. On this basis, a precise diagnostic model and an intervention model for this specialty and disease are constructed.

2. The method for constructing a precise traditional Chinese medicine diagnosis and treatment model based on real clinical data according to claim 1, wherein: During the construction of the diagnostic model, it is necessary to determine the staging information of each disease and the grouped diagnostic information of each syndrome.

3. A method for constructing a precise traditional Chinese medicine diagnosis and treatment model based on real clinical data according to claim 2, characterized in that: The construction of the intervention model is based on precise diagnosis to determine the core prescriptions and the addition and subtraction of medications for each syndrome group.

4. A method for constructing a precise traditional Chinese medicine diagnosis and treatment model based on real clinical data according to claim 1, wherein It also includes the following standardized operation process steps for constructing the diagnosis and treatment model: Step 1: Disease identification: After completing the collection of real clinical data and related standards for this specific disease, classify and code the disease name according to the national standard "Classification and Codes of Traditional Chinese Medicine Syndromes and Diseases" (GB / T 15657-2021). Step 2: Disease staging: After determining the disease name, customize the disease staging principles and methods according to the research design of the disease being studied and the actual clinical situation. Step 3: Syndrome differentiation: In accordance with the requirements and methods of traditional Chinese medicine clinical syndrome differentiation diagnosis, identify the syndrome to which the research belongs through comprehensive analysis of syndrome differentiation elements such as clinical symptoms, tongue manifestations, pulse conditions, and objective examination results. Step 4: Syndrome grouping: Obtain the core prescription information from the clinical data. If there is no core prescription information, it is determined through clustering analysis of medication data, and at the same time, syndrome grouping is carried out based on the usage of the core prescription. Step 5: Determine syndrome differentiation elements: After disease staging and syndrome grouping, determine the syndrome differentiation elements of each syndrome group and establish a syndrome differentiation element database according to the syndrome grouping. Step 6: Identify concurrent diseases, syndromes, and symptoms: Based on the syndrome differentiation element database of each syndrome group, identify and extract the information of concurrent diseases, syndromes, and symptoms of each syndrome group, and establish a database for concurrent diseases, syndromes, and symptoms of this syndrome group according to the three categories of concurrent diseases, concurrent syndromes, and concurrent symptoms. Step 7: Determine the medications of the core prescription. Step 8: Determine the addition and subtraction of medications for concurrent diseases, syndromes, and symptoms.

5. A method for constructing a precise traditional Chinese medicine diagnosis and treatment model based on real clinical data according to claim 4, characterized in that: In Step 1 mentioned above, if the disease name in the original data conforms to the coding specification of "Classification and Codes of Traditional Chinese Medicine Syndromes and Diseases", it is coded based on it as the disease name.

6. The method for constructing a precise traditional Chinese medicine diagnosis and treatment model based on real clinical data according to claim 5, characterized in that: If the disease in the original data is expressed in a way that does not conform to the coding specification of "Classification and Codes of Traditional Chinese Medicine Syndromes and Diseases", it is reclassified and coded according to the clinical reality and expert consultation; if a Western medicine disease name is used, the disease name should be determined and coded according to the disease classification of ICD.

7. A method for constructing a precise traditional Chinese medicine diagnosis and treatment model based on real clinical data according to claim 4, characterized in that: In Step 3 mentioned above, determine the syndrome classification and codes of the disease being studied according to the national standard "Classification and Codes of Traditional Chinese Medicine Syndromes and Diseases".

8. A method for constructing a precise traditional Chinese medicine diagnosis and treatment model based on real clinical data according to claim 4, characterized in that: In Step 4 mentioned above, if there is no core prescription information in the clinical data, clustering analysis is performed on its medication data to determine the core prescription for each disease syndrome, and the syndrome is grouped according to its different core prescriptions, and a syndrome grouping information database for this disease syndrome is established.

9. A method for constructing a precise traditional Chinese medicine diagnosis and treatment model based on real clinical data according to claim 4, characterized in that: Step 7 is used to determine the drug names and dosage information of each core prescription, and establish a core prescription medication information database for this disease syndrome according to the syndrome grouping.

10. A method for constructing a precise traditional Chinese medicine diagnosis and treatment model based on real clinical data according to claim 4, characterized in that: In Step 8, the core prescription medications in the clinical prescription data are deleted to obtain the medication information corresponding to concurrent complex diseases, syndromes, and symptoms, and then, based on the specifications of the functions and indications of the medications, the medications corresponding to each of the concurrent complex diseases, syndromes, and symptoms are determined.

Images