Traditional Chinese medicine constitution identification method and device for coronary heart disease patient and computer readable storage medium

Abstract

The invention discloses a traditional Chinese medicine constitution identification method and device for a coronary heart disease patient and a computer readable storage medium. The traditional Chinese medicine constitution identification method comprises the following steps: acquiring traditional Chinese medicine constitution identification data and a disease factor of a patient; performing classification based on the disease factors and the traditional Chinese medicine constitution recognition data, and determining the traditional Chinese medicine constitution types of the patients of the corresponding disease factors; fusing the disease factors, the traditional Chinese medicine constitution recognition information and the traditional Chinese medicine constitution types of the patients of the corresponding disease factors; training an LDA topic model through data obtained after fusion to obtain a traditional Chinese medicine constitution identification model; and inputting the target disease factor of the to-be-identified patient into the traditional Chinese medicine constitution identification model for identification to obtain a target traditional Chinese medicine constitution type of the to-be-identified patient. The technical problem that the correlation between the coronary heart disease and the traditional Chinese medicine constitution is not considered, so that the intelligent analysis of the coronary heart disease and the traditional Chinese medicine constitution is not accurate enough is solved.

Description

Technical Field

[0001] This application relates to the field of health monitoring, and more specifically, to a method, device, and computer-readable storage medium for identifying the constitution of patients with coronary heart disease using traditional Chinese medicine. Background Technology

[0002] In the prior art, patent CN116364300A provides a method for identifying TCM constitution, including: completing punctuation in doctor-patient question-and-answer data; performing role identification, text and basic information splicing on the punctuated doctor-patient question-and-answer data to obtain complete continuous text; classifying the complete continuous text using a pre-trained Roberta model to obtain a first constitution score; identifying symptom entities in the doctor-patient question-and-answer data; matching the symptom entities with a preset symptom standard scale to obtain direct and indirect symptoms; merging direct and indirect symptoms to obtain the target direct symptom; calculating a second constitution score based on the patient's basic information and the target direct symptom using prediction scale rules; combining the first and second constitution scores to obtain a comprehensive constitution score; and taking the highest score in the comprehensive constitution score as the patient's constitution.

[0003] This patent mainly determines the TCM constitution score based on doctor-patient Q&A data, and determines the TCM constitution score based on symptoms and basic patient information. Finally, it combines the two scores to determine the TCM constitution. However, it does not consider the correlation between coronary heart disease and TCM constitution, which leads to the inaccuracy of the intelligent analysis of TCM constitution for coronary heart disease.

[0004] There is currently no effective solution to the problem that the related technologies do not consider the correlation between coronary heart disease and TCM constitution, resulting in inaccurate intelligent analysis of TCM constitution in coronary heart disease. Summary of the Invention

[0005] The main purpose of this application is to provide a method, device and computer-readable storage medium for identifying the TCM constitution of patients with coronary heart disease, so as to solve the problem that the intelligent analysis of TCM constitution of coronary heart disease is not accurate enough because the correlation between coronary heart disease and TCM constitution is not considered.

[0006] To achieve the above objectives, according to one aspect of this application, a method for identifying the constitution of patients with coronary heart disease using traditional Chinese medicine is provided.

[0007] The method for identifying the TCM constitution of patients with coronary heart disease according to this application includes: acquiring TCM constitution identification data and the patient's disease factors; classifying the disease factors and TCM constitution identification data to determine the TCM constitution type of the patient with the corresponding disease factors; fusing the disease factors, TCM constitution identification information, and the TCM constitution type of the patient with the corresponding disease factors; training an LDA topic model with the fused data to obtain a TCM constitution identification model; and inputting the target disease factors of the patient to be identified into the TCM constitution identification model for identification to obtain the target TCM constitution type of the patient to be identified.

[0008] Furthermore, based on the disease factors and TCM constitution identification data, classification is performed to determine the TCM constitution type of patients with the corresponding disease factors, including: Disease factors were used as keywords to search for TCM constitution classification and judgment criteria and the nine-point TCM constitution identification scale in TCM constitution identification data. When the same keyword is found, the TCM constitution label corresponding to that keyword is taken as the patient's TCM constitution type.

[0009] Furthermore, the disease factors, TCM constitution identification information, and the TCM constitution type of the corresponding disease factor patient are integrated, including: S301. Scan and obtain all disease factors, TCM constitution identification information, and TCM constitution type of patients with corresponding disease factors, and generate candidate k-itemsets; S302. Count each item in the candidate k-item set and calculate its support. Determine the preset minimum support and delete items in the candidate k-item set whose support is less than the preset minimum support to obtain the frequent k-item set. If the frequent k-itemset is empty, then the set of frequent k-1 itemsets is returned directly as the result of the algorithm, and the algorithm ends. S303. If the frequent k-itemsets are not empty, generate candidate k+1 itemsets by connecting them. Repeat step S302 until the frequent k-itemsets are empty. The algorithm ends and the result is used as the fused data set.

[0010] Furthermore, determine the preset minimum support level, including: Determine the selection method for the preset minimum support, select several preset minimum support values, establish a comprehensive scoring formula, obtain the score value for each preset minimum support value, and select the preset minimum support value with the highest score value.

[0011] Furthermore, an LDA topic model is trained using the fused data to obtain a TCM constitution identification model, including: constructing an initial LDA topic model; extracting keywords from the fused dataset using the TF-IDF and TextRank algorithms to form an indicator database; randomly assigning a topic to each indicator data; and referring to all the topics of the indicator data as subtopics. The index database is scanned, and the improved Gibbs sampling method is used to sample each index data until convergence, so as to obtain the optimal hyperparameters α and β. The optimal hyperparameters are substituted into the LDA topic model, and the fused dataset is used for retraining to obtain the trained LDA topic model. At the same time, the index data-topic matrix and the topic-index data matrix are obtained.

[0012] Furthermore, keywords are extracted from the fused dataset using the TF-IDF and TextRank algorithms to construct an indicator database, including: All data in the fused dataset are divided into N partition sets according to preset biochemical indicator categories; the first weight of each partition set in the fused dataset is calculated based on the TF-IDF algorithm formula. A network node model is constructed based on N partition sets, and the TextRank function is used to obtain the second weight of each partition set in all the fused datasets. Based on the first weight in descending order, select the first m weight corresponding to the partition set as the first candidate word set for keywords; According to the second weight in descending order, select the partition set corresponding to the first m weights as the second candidate word set of keywords; The partition sets in the first and second candidate word sets are matched. If a partition set exists in both the first and second candidate word sets, then the partition set is added to the final candidate set, and the final candidate set is used as the indicator database.

[0013] Furthermore, an improved Gibbs sampling method was used to sample data for each indicator, including: The original Gibbs sampling method is optimized based on Monte Carlo sampling. The calculation formula for the improved Gibbs sampling method is as follows: in, Let z represent the topic probability distribution of the feature word d at position i in the text data. i This represents the topic corresponding to the feature word at the i-th position in the text data. This represents the total number of feature words for the k-th topic in the m-th document. This represents the number of the t-th feature word in the k-th topic. This indicates the removal of the feature word with index i. The total number of feature words is D, d, j, t = 1, 2, L, D. weight(j), weight(d), and weight(t) all represent the weights of the feature words, and α and β are hyperparameters.

[0014] Furthermore, the optimal hyperparameters are substituted into the LDA topic model, and the fused dataset is used for retraining to obtain a trained LDA topic model. After obtaining the index data-topic matrix and the topic-index data matrix, the following are also included: The frequency of statistical indicator data-topic matrix and topic-indicator data matrix is ​​used as topic intensity. The average topic intensity value is obtained by summing and averaging the intensities of subtopics. The ratio of topic intensity to average topic intensity (TIP) is calculated. TIP is compared with a preset threshold to determine the topic type of TCM constitution classification and categorization of coronary heart disease.

[0015] To achieve the above objectives, according to another aspect of this application, a device for identifying the constitution of patients with coronary heart disease using traditional Chinese medicine is provided.

[0016] The TCM constitution identification device for coronary heart disease patients according to this application includes: an acquisition module for acquiring TCM constitution identification data and the patient's disease factors; a classification module for classifying based on the disease factors and TCM constitution identification data to determine the TCM constitution type of the patient with the corresponding disease factor; a fusion module for fusing the disease factors, TCM constitution identification information, and the TCM constitution type of the patient with the corresponding disease factor; a training module for training an LDA topic model using the fused data to obtain a TCM constitution identification model; and an identification module for inputting the target disease factor of the patient to be identified into the TCM constitution identification model for identification to obtain the target TCM constitution type of the patient to be identified.

[0017] To achieve the above objectives, according to another aspect of this application, a computer-readable storage medium is provided.

[0018] According to the computer-readable storage medium of this application, the computer-readable storage medium stores a computer program, wherein the computer program is configured to execute the traditional Chinese medicine constitution identification device for patients with coronary heart disease when it is run.

[0019] In this embodiment, a method for identifying the TCM constitution of patients with coronary heart disease is adopted. This involves acquiring TCM constitution identification data and the patient's disease factors; classifying the disease factors and TCM constitution identification data to determine the TCM constitution type of the patient with the corresponding disease factor; fusing the disease factors, TCM constitution identification information, and the TCM constitution type of the patient with the corresponding disease factor; training an LDA topic model with the fused data to obtain a TCM constitution identification model; and inputting the target disease factor of the patient to be identified into the TCM constitution identification model for identification to obtain the target TCM constitution type of the patient. This method fully considers the correlation between coronary heart disease and TCM constitution, thereby effectively improving the accuracy of intelligent analysis of TCM constitution in coronary heart disease and solving the technical problem that the intelligent analysis of TCM constitution in coronary heart disease is not accurate enough due to the lack of consideration of the correlation between coronary heart disease and TCM constitution. Attached Figure Description

[0020] The accompanying drawings, which form part of this application, are used to provide a further understanding of the application and to make other features, objects, and advantages of the application more apparent. The illustrative embodiments and descriptions of this application are used to explain the application and do not constitute an undue limitation of the application. In the drawings: Figure 1 This is a flowchart illustrating the method for identifying the constitution of patients with coronary heart disease according to an embodiment of this application. Figure 2 This is a schematic diagram of the structure of a TCM constitution identification device for patients with coronary heart disease according to an embodiment of this application. Detailed Implementation

[0021] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0022] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this application described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0023] In this application, the terms "upper," "lower," "left," "right," "front," "rear," "top," "bottom," "inner," "outer," "middle," "vertical," "horizontal," "lateral," and "longitudinal" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are primarily for the purpose of better describing the invention and its embodiments, and are not intended to limit the indicated device, element, or component to having a specific orientation, or to be constructed and operated in a specific orientation.

[0024] Furthermore, in addition to indicating direction or positional relationship, some of the aforementioned terms may also have other meanings. For example, the term "above" may also be used in certain situations to indicate a dependency or connection. Those skilled in the art can understand the specific meaning of these terms in this invention based on the specific circumstances.

[0025] Furthermore, the terms "installation," "setup," "equipped with," "connection," "linking," and "socketing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral structure; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium, or an internal connection between two devices, components, or parts. Those skilled in the art can understand the specific meaning of these terms in this invention based on the specific circumstances.

[0026] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0027] According to embodiments of the present invention, a method for identifying the constitution of patients with coronary heart disease using traditional Chinese medicine is provided, such as... Figure 1 As shown, the method includes the following steps S101 to S105: Step S101: Obtain TCM constitution identification data and the patient's disease factors; Traditional Chinese medicine (TCM) constitution identification data refers to the evaluation criteria used to identify TCM constitutions. Patient disease factors refer to factors related to coronary heart disease in patients with coronary heart disease. In this invention, the evaluation criteria are the TCM constitution classification and judgment standards (ZYYXH / T1572009) promulgated by the China Association of Traditional Chinese Medicine in 2009, combined with the 9-point TCM constitution identification scale compiled by Wang Qi.

[0028] It is important to understand that the constitution types can be divided into nine subscales, including balanced constitution, qi deficiency constitution, yang deficiency constitution, yin deficiency constitution, phlegm-dampness constitution, damp-heat constitution, blood stasis constitution, qi stagnation constitution, and special constitution, with a total of 67 items of constitution information.

[0029] It is also necessary to understand that patient factors include, but are not limited to, patient-related information such as age, gender, weight, and family history of genetic diseases.

[0030] Step S102: Based on the disease factors and TCM constitution identification data, classify and determine the TCM constitution type of patients with the corresponding disease factors; The criteria for judging TCM constitution involve information such as age, gender, weight, and genetic history. In other words, this information is related to the pathogenic factors of patients with coronary heart disease. Therefore, based on this relationship, TCM constitution types and pathogenic factors can be matched through human identification or intelligent search and comparison technology, thereby determining the pathogenic factors corresponding to each TCM constitution type.

[0031] In this embodiment, preferably, the TCM constitution type of the patient is determined by classifying the disease factors and TCM constitution identification data, including: using the disease factors as keywords to search in the TCM constitution classification and judgment criteria and the nine-point TCM constitution identification scale of the TCM constitution identification data; when the same keyword is found, the TCM constitution label corresponding to the keyword is taken as the patient's TCM constitution type.

[0032] This study classifies the pathogenic factors of coronary heart disease according to TCM constitution types, thus enabling the preliminary establishment of a correlation between coronary heart disease and TCM constitution. In addition, the use of keyword search and matching can effectively improve the efficiency of establishing the correlation between coronary heart disease and TCM constitution, thereby reducing the manpower cost of labeling datasets.

[0033] It is worth noting that there is a certain correlation between age, gender, and weight in patients with coronary heart disease. For example, Qi deficiency, blood stasis, phlegm-dampness, and Yang deficiency are generally the main constitution types in people with coronary heart disease, and the constitution type is related to the degree of coronary artery stenosis. Patients with Qi deficiency, blood stasis, and phlegm-dampness constitutions tend to have more severe coronary artery stenosis.

[0034] In addition, the most common constitution types among male patients with coronary heart disease (≥30%) are phlegm-dampness, qi deficiency, and yang deficiency, while the most common constitution type among female patients (≥30%) is phlegm-dampness. There is a large age difference among patients with different TCM constitutions, especially among patients with coronary heart disease who have qi stagnation or yang deficiency constitutions, the age of the patients is relatively old.

[0035] For example, the distribution of TCM constitution types differs among coronary heart disease patients of different weights. Here, referring to the classification standard of BMI for Chinese adults by the China Obesity Working Group, weight is divided into four groups: underweight (BMI < 18.5 kg / m²), normal weight (18.5 ≤ BMI ≤ 23.9 kg / m²), overweight (24 kg / m² ≤ BMI ≤ 27.9 kg / m²), and obese (BMI ≥ 28 kg / m²). Comparison revealed that patients with phlegm-dampness constitution have a higher BMI among coronary heart disease patients. Among patients of different weights, those with phlegm-dampness constitution have the highest BMI.

[0036] Step S103: Integrate the disease factor, TCM constitution identification information, and the TCM constitution type of the patient with the corresponding disease factor; In this invention, the Apriori algorithm is used to fuse the disease factors, TCM constitution identification information, and the TCM constitution type of patients with the corresponding disease factors to generate a fused dataset. The fused dataset can further reflect the correlation between coronary heart disease and TCM constitution, and the accuracy of the correlation between the data is also higher, thus further improving the accuracy of TCM constitution analysis.

[0037] It's important to understand that the basic idea of ​​the Apriori algorithm is based on the concept of frequent itemsets, which are sets of itemsets that appear more than a preset threshold in a dataset. It iteratively generates candidate sets of different lengths and leverages the property of frequent itemsets (if an itemset is frequent, then all its non-empty subsets are also frequent) to reduce the search space. The algorithm mainly consists of three steps: data processing, frequent itemset generation, and association rule generation. When generating frequent itemsets, the algorithm identifies all possible frequent itemsets (i.e., candidate itemsets) and then selects the most frequent itemsets based on minimum support counts. When generating association rules, the algorithm calculates frequent itemsets and rules based on minimum support and confidence levels.

[0038] Specifically, the disease factors, TCM constitution identification information, and the TCM constitution type of the patient with the corresponding disease factors are integrated, including: S301. Scan and obtain all disease factors, TCM constitution identification information, and TCM constitution type of patients with corresponding disease factors, and generate candidate k-itemsets; S302. Count each item in the candidate k-item set and calculate its support. Determine the preset minimum support and delete items in the candidate k-item set whose support is less than the preset minimum support to obtain the frequent k-item set. If the frequent k-itemset is empty, then the set of frequent k-1 itemsets is returned directly as the result of the algorithm, and the algorithm ends. S303. If the frequent k-itemsets are not empty, generate candidate k+1 itemsets by connecting them. Repeat step S302 until the frequent k-itemsets are empty. The algorithm ends and the result is used as the fused data set.

[0039] Further, determining the preset minimum support includes: determining the selection method for the preset minimum support, selecting several preset minimum support values, establishing a comprehensive scoring formula, obtaining the score value of each preset minimum support value, and selecting the preset minimum support value with the highest score value.

[0040] Step S104: Train the LDA topic model using the fused data to obtain the TCM constitution identification model; By training the LDA topic model with the classified and fused data, a TCM constitution identification model that can identify TCM constitution types can be obtained.

[0041] It's important to understand that LDA (Latent Dirichlet Allocation) is a document topic generation model, also known as a three-layer Bayesian probabilistic model, comprising a three-layer structure of words, topics, and documents. As a generational model, it means that we believe each word in an article is obtained through a process of "selecting a certain topic with a certain probability, and then selecting a word from that topic with a certain probability." The document-to-topic distribution follows a multinomial distribution, and the topic-to-word distribution also follows a multinomial distribution.

[0042] It's also important to understand that LDA is an unsupervised machine learning technique used to identify latent topical information in large document collections or corpora. It employs the bag-of-words approach, treating each document as a word frequency vector, thus transforming textual information into easily modelable numerical information. However, the bag-of-words method doesn't consider the order of words, which simplifies the problem but also provides an opportunity for model improvement. Each document represents a probability distribution of some topics, and each topic represents a probability distribution of many words.

[0043] Specifically, the LDA topic model is trained using the fused data to obtain a TCM constitution identification model, including: constructing an initial LDA topic model, extracting keywords from the fused dataset using the TF-IDF and TextRank algorithms to form an indicator database, randomly assigning a topic to each indicator data, and the topics of all indicator data are called subtopics; The index database is scanned, and the improved Gibbs sampling method is used to sample each index data until convergence, so as to obtain the optimal hyperparameters α and β. The optimal hyperparameters are substituted into the LDA topic model, and the fused dataset is used for retraining to obtain the trained LDA topic model. At the same time, the index data-topic matrix and the topic-index data matrix are obtained.

[0044] Furthermore, keywords are extracted from the fused dataset using the TF-IDF and TextRank algorithms to construct the indicator database, which specifically includes: All data in the fused dataset are divided according to preset biochemical indicator categories to obtain N partition sets; The first weight of each partition set in the fused dataset is calculated based on the TF-IDF algorithm formula; A network node model is constructed based on N partition sets, and the TextRank function is used to obtain the second weight of each partition set in all fused datasets. Based on the first weight in descending order, select the first m weight corresponding to the partition set as the first candidate word set for keywords; According to the second weight in descending order, select the partition set corresponding to the first m weights as the second candidate word set of keywords; The partition sets in the first and second candidate word sets are matched. If a partition set exists in both the first and second candidate word sets, then the partition set is added to the final candidate set, and the final candidate set is used as the indicator database.

[0045] Furthermore, an improved Gibbs sampling method is used to sample data for each indicator, including: The original Gibbs sampling method is optimized based on Monte Carlo sampling. The calculation formula for the improved Gibbs sampling method is as follows: in, Let z represent the topic probability distribution of the feature word d at position i in the text data. i This represents the topic corresponding to the feature word at the i-th position in the text data. This represents the total number of feature words for the k-th topic in the m-th document. This represents the number of the t-th feature word in the k-th topic. This indicates the removal of the feature word with index i. The total number of feature words is D, d, j, t = 1, 2, L, D. weight(j), weight(d), and weight(t) all represent the weights of the feature words, and α and β are hyperparameters.

[0046] Furthermore, the optimal hyperparameters are substituted into the LDA topic model, and the fused dataset is used for retraining to obtain a trained LDA topic model. After obtaining the indicator data-topic matrix and the topic-indicator data matrix, the following are also included: The frequency of statistical indicator data-topic matrix and topic-indicator data matrix is ​​used as topic intensity. The average topic intensity value is obtained by summing and averaging the intensities of subtopics. The ratio of topic intensity to average topic intensity (TIP) is calculated. TIP is compared with a preset threshold to determine the topic type of TCM constitution classification and categorization of coronary heart disease.

[0047] Step S105: Input the target disease factor of the patient to be identified into the TCM constitution identification model for identification, and obtain the target TCM constitution type of the patient to be identified.

[0048] Patients seeking identification only need to input their target disease factors—age, gender, weight, and family history of genetic diseases—into a computer. The computer, with the corresponding program stored, can then use a pre-trained TCM constitution identification model to identify the patient's target TCM constitution type based on these factors. This effectively improves the accuracy of intelligent TCM constitution analysis for coronary heart disease.

[0049] As can be seen from the above description, the present invention achieves the following technical effects: In this embodiment, a method for identifying the TCM constitution of patients with coronary heart disease is adopted. This involves acquiring TCM constitution identification data and the patient's disease factors; classifying the disease factors and TCM constitution identification data to determine the TCM constitution type of the patient with the corresponding disease factor; fusing the disease factors, TCM constitution identification information, and the TCM constitution type of the patient with the corresponding disease factor; training an LDA topic model with the fused data to obtain a TCM constitution identification model; and inputting the target disease factor of the patient to be identified into the TCM constitution identification model for identification to obtain the target TCM constitution type of the patient. This method fully considers the correlation between coronary heart disease and TCM constitution, thereby effectively improving the accuracy of intelligent analysis of TCM constitution in coronary heart disease and solving the technical problem that the intelligent analysis of TCM constitution in coronary heart disease is not accurate enough due to the lack of consideration of the correlation between coronary heart disease and TCM constitution.

[0050] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0051] According to embodiments of the present invention, an apparatus is also provided for implementing the above-described method for identifying the constitution of patients with coronary heart disease using traditional Chinese medicine, such as... Figure 2 As shown, the device includes: an acquisition module 10 for acquiring TCM constitution identification data and the patient's disease factors; a classification module 20 for classifying based on the disease factors and TCM constitution identification data to determine the TCM constitution type of the patient with the corresponding disease factor; a fusion module 30 for fusing the disease factors, TCM constitution identification information, and the TCM constitution type of the patient with the corresponding disease factor; a training module 40 for training an LDA topic model using the fused data to obtain a TCM constitution identification model; and an identification module 50 for inputting the target disease factor of the patient to be identified into the TCM constitution identification model for identification to obtain the target TCM constitution type of the patient to be identified. This device achieves the same technical effect as the TCM constitution identification method for patients with coronary heart disease.

[0052] Obviously, those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a storage device for execution by a computing device, or fabricating them separately as individual integrated circuit modules, or fabricating multiple modules or steps as a single integrated circuit module. Thus, the present invention is not limited to any particular hardware and software combination.

[0053] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for identifying the constitution of patients with coronary heart disease using Traditional Chinese Medicine, characterized in that, include: To obtain TCM constitution identification data and patients' disease factors; Based on the disease factors and TCM constitution identification data, the TCM constitution type of patients with the corresponding disease factors is determined. The disease factors, TCM constitution identification information, and TCM constitution types of patients with corresponding disease factors are integrated; The LDA topic model was trained using the fused data to obtain a TCM constitution identification model; The target disease factors of the patient to be identified are input into the TCM constitution identification model for identification, and the target TCM constitution type of the patient to be identified is obtained.

2. The method for identifying traditional Chinese medicine constitution according to claim 1, characterized in that, Based on the aforementioned disease factors and TCM constitution identification data, the TCM constitution type of patients with corresponding disease factors is determined, including: Disease factors were used as keywords to search for TCM constitution classification and judgment criteria and the nine-point TCM constitution identification scale in TCM constitution identification data. When the same keyword is found, the TCM constitution label corresponding to that keyword is taken as the patient's TCM constitution type.

3. The method for identifying traditional Chinese medicine constitution according to claim 1, characterized in that, The disease factors, TCM constitution identification information, and the TCM constitution type of patients with the corresponding disease factors are integrated, including: S301. Scan and obtain all disease factors, TCM constitution identification information, and TCM constitution type of patients with corresponding disease factors, and generate candidate k-itemsets; S302. Count each item in the candidate k-item set and calculate its support. Determine the preset minimum support and delete items in the candidate k-item set whose support is less than the preset minimum support to obtain the frequent k-item set. If the frequent k-itemset is empty, then the set of frequent k-1 itemsets is returned directly as the result of the algorithm, and the algorithm ends. S303. If the frequent k-itemsets are not empty, generate candidate k+1 itemsets by connecting them. Repeat step S302 until the frequent k-itemsets are empty. The algorithm ends and the result is used as the fused data set.

4. The method for identifying traditional Chinese medicine constitution according to claim 3, characterized in that, Determining the preset minimum support includes: determining the selection method for the preset minimum support, selecting several preset minimum support values, establishing a comprehensive scoring formula, obtaining the score value for each preset minimum support value, and selecting the preset minimum support value with the highest score value.

5. The method for identifying traditional Chinese medicine constitution according to claim 1, characterized in that, By training an LDA topic model with the fused data, a Traditional Chinese Medicine (TCM) constitution identification model is obtained, including: An initial LDA topic model is constructed, and keywords are extracted from the fused dataset using the TF-IDF and TextRank algorithms to form an indicator database. Each indicator data is randomly assigned a topic, and the topics of all indicator data are called subtopics. The index database is scanned, and the improved Gibbs sampling method is used to sample each index data until convergence, so as to obtain the optimal hyperparameters α and β. The optimal hyperparameters are substituted into the LDA topic model, and the fused dataset is used for retraining to obtain the trained LDA topic model. At the same time, the index data-topic matrix and the topic-index data matrix are obtained.

6. The method for identifying traditional Chinese medicine constitution according to claim 5, characterized in that, Keywords are extracted from the fused dataset using the TF-IDF and TextRank algorithms to form an indicator database, including: All data in the fused dataset are divided into N partition sets according to preset biochemical indicator categories; the first weight of each partition set in the fused dataset is calculated based on the TF-IDF algorithm formula. A network node model is constructed based on N partition sets, and the TextRank function is used to obtain the second weight of each partition set in all the fused datasets. Based on the first weight in descending order, select the first m weight corresponding to the partition set as the first candidate word set for keywords; According to the second weight in descending order, select the partition set corresponding to the first m weights as the second candidate word set of keywords; The partition sets in the first and second candidate word sets are matched. If a partition set exists in both the first and second candidate word sets, then the partition set is added to the final candidate set, and the final candidate set is used as the indicator database.

7. The method for identifying traditional Chinese medicine constitution according to claim 5, characterized in that, The improved Gibbs sampling method was used to sample data for each indicator, including: The original Gibbs sampling method is optimized based on Monte Carlo sampling. The calculation formula for the improved Gibbs sampling method is as follows: in, Let z represent the topic probability distribution of the feature word d at position i in the text data. i This represents the topic corresponding to the feature word at the i-th position in the text data. This represents the total number of feature words for the k-th topic in the m-th document. This represents the number of the t-th feature word in the k-th topic. This indicates the removal of the feature word with index i. The total number of feature words is D, where d,j,t = 1, 2, L, D. weight(j), weight(d), and weight(t) all represent the weights of the feature words, and α and β are hyperparameters.

8. The method for identifying traditional Chinese medicine constitution according to claim 5, characterized in that, Substituting the optimal hyperparameters into the LDA topic model, and retraining using the fused dataset, yields a trained LDA topic model. This, along with the resulting index-topic matrix and topic-index matrix, also includes: The frequency of statistical indicator data-topic matrix and topic-indicator data matrix is ​​used as topic intensity. The average topic intensity value is obtained by summing and averaging the intensities of subtopics. The ratio of topic intensity to average topic intensity (TIP) is calculated. TIP is compared with a preset threshold to determine the topic type of TCM constitution classification and categorization of coronary heart disease.

9. A device for identifying the constitution of patients with coronary heart disease using traditional Chinese medicine, characterized in that, include: The acquisition module is used to acquire TCM constitution identification data and patients' disease factors; The classification module is used to classify patients based on the disease factors and TCM constitution identification data, and to determine the TCM constitution type of patients with the corresponding disease factors. The fusion module is used to fuse the disease factors, TCM constitution identification information, and the TCM constitution type of the patient with the corresponding disease factors; The training module is used to train the LDA topic model using the fused data to obtain the TCM constitution identification model. The identification module is used to input the target disease factors of the patient to be identified into the TCM constitution identification model for identification, and to obtain the target TCM constitution type of the patient to be identified.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein the computer program is configured to execute, when running, the traditional Chinese medicine constitution identification device for patients with coronary heart disease as described in any one of claims 1 to 8.

Images