Traditional Chinese medicine constitution typing method and system based on obesity with disease data

Abstract

The invention discloses a traditional Chinese medicine constitution typing method and system based on obesity with disease data, and the method comprises the steps: collecting the disease condition data information of obese patients, and carrying out the desensitization of the data information; carrying out standardization processing on the disease condition data information and preset identification feature data of nine constitutions of traditional Chinese medicine, and marking the types of the constitutions of the traditional Chinese medicine with obesity accompanying diseases; screening the characteristic data corresponding to the traditional Chinese medicine constitution of the obesity accompanying disease based on an IV value analysis method to obtain characteristic data for model construction, and forming a data set; based on a logistic regression algorithm, using the data set to construct a traditional Chinese medicine constitution typing model fused with obese patient combined disease data; and predicting the accompanying disease sample data of the to-be-tested patient through the traditional Chinese medicine constitution typing model to obtain the obese traditional Chinese medicine constitution category. The reliability and typing precision of data analysis are improved, and traditional Chinese medicine precise prevention of obesity and complications thereof is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent identification of TCM constitutions, and in particular to a TCM constitution typing method and system based on obesity-associated disease data. Background Art

[0002] At present, obesity is a major global public health problem, and the research on the correlation between its complications such as metabolic syndrome and cardiovascular disease and traditional Chinese medicine constitution has become a hot topic. While existing Traditional Chinese Medicine (TCM) constitution typing methods have achieved initial success in people with simple obesity, they face significant technical bottlenecks in the field of obesity-associated diseases. Traditional studies have failed to analyze the constitutional differences among obese individuals with specific diseases such as diabetes, hypertension, coronary heart disease, hyperuricemia, and gout, ignoring the specific associations between the pathological mechanisms of different diseases and constitutional biases. For example, patients with diabetes have a higher proportion of people with qi deficiency, and patients with coronary heart disease have significant blood stasis characteristics, resulting in a lack of disease-specific constitutional typing. Furthermore, obesity-associated disease data are multi-source and complex, containing disease status, indicator data, and demographic information. Existing methods lack systematic and standardized data desensitization and standardization, such as converting qualitative variables to numerical values, dividing continuous data into intervals, imputing missing values ​​with the mean, and screening key features. This makes it difficult to effectively eliminate noise interference, reducing model training efficiency and accuracy. Furthermore, traditional typing often relies on empirical classification or simple statistical methods, without introducing machine learning algorithms for data-driven modeling. In particular, there is a lack of logistic regression models that integrate disease data with constitutional characteristics, making it impossible to quantify the impact of disease characteristics on constitutional type. Therefore, how to integrate the combined disease data of obese patients, construct a classification model that takes into account both disease specificity and overall constitution, and realize data standardization, scientific feature screening and precise modeling is a basic problem to be solved in intelligent and precise TCM constitution classification. Summary of the Invention

[0003] In view of this, the present invention proposes a TCM constitution classification method and system based on obesity-associated disease data, which can quickly and accurately predict patients and determine their obesity TCM constitution category. The present invention provides the following technical solutions: A method for TCM constitution typing based on obesity-associated disease data, comprising: Collecting disease status data and information of obese patients and performing desensitization processing on the data; Standardize the disease condition data and the preset nine TCM constitution identification feature data, and mark the TCM constitution categories of obesity-associated diseases; Based on the IV value analysis method, the characteristic data corresponding to the TCM constitution of obesity-associated diseases were screened to obtain the characteristic data for model construction and form a data set; Based on the logistic regression algorithm, the data set is used to construct a TCM constitution classification model that integrates the combined disease data of obese patients; the TCM constitution classification model is used to predict the concomitant disease sample data of the tested patients to obtain the TCM constitution category of obesity.

[0004] Optionally, collecting disease status data information of obese patients and performing desensitization processing on the data information includes: Collect data on diabetes, hypertension, coronary heart disease, hyperuricemia, gout and other conditions and related indicators of obese patients, as well as basic information such as gender, age, weight, and occupation of the patients; Perform data cleaning on the collected diagnostic test data, including converting qualitative variables into numerical variables, processing outliers and missing values, and deleting duplicate values; Add classification labels to the cleaned detection data based on the diagnostic data to build a dataset that can be used by the model.

[0005] Optionally, the disease condition data information and the preset TCM nine constitution identification feature data are standardized, and the TCM constitution categories of obesity-associated diseases are marked, including: Type conversion, data filling and data deletion operations are performed on disease status data information and the nine constitution identification feature data of traditional Chinese medicine; the traditional Chinese medicine constitution types of obesity-associated diseases are marked as a single category, and the category labels correspond to the following traditional Chinese medicine constitution categories: balanced constitution, qi deficiency constitution, yang deficiency constitution, yin deficiency constitution, phlegm-damp constitution, damp-heat constitution, blood stasis constitution, qi stagnation constitution, and special constitution.

[0006] Optionally, the IV value analysis method is used to screen the characteristic data corresponding to the TCM constitution of obesity-associated diseases to obtain characteristic data for model construction, forming a data set, including: The IV value analysis method was used to measure the influence of each feature on the TCM body type of obesity, and the correlation degree was calculated by comparing the ratio of black and white samples hit by the feature with the ratio of total black and white samples; Feature data that are consistent with the prediction of the TCM constitution category for obesity are screened out, including the disease conditions and indicator data of diabetes, hypertension, coronary heart disease, hyperuricemia, and gout, and a data set for model construction is formed based on the screened feature data.

[0007] Optionally, the method of constructing a TCM constitution classification model integrating the combined disease data of obese patients using the data set based on a logistic regression algorithm includes: Based on linear regression, the linear combination result of the characteristic data of the data set is mapped to the interval (0, 1) through the Sigmoid function, and the probability of predicting the obesity TCM constitution type is calculated; The maximum likelihood estimation is used to construct the loss function, the logarithm of the likelihood function is taken and converted into a minimized negative log-likelihood function, and the optimal coefficient vector is iteratively solved by the gradient descent method to make the loss function converge, and finally a TCM constitution classification model is constructed.

[0008] Optionally, the TCM constitution classification model is used to predict the concomitant disease sample data of the patient to be tested, and the obesity TCM constitution category obtained includes: Standardizing the sample data of the patients with concomitant diseases to be tested, and converting the sample data into an input vector consistent with the format of the feature data in the dataset; The input vector is input into the logistic regression model to calculate the probability value of the patient to be tested belonging to each TCM constitution category; based on a preset probability threshold, the constitution category with the highest probability value is determined as the obese TCM constitution category of the patient to be tested.

[0009] The present invention further discloses a TCM constitution classification system based on obesity-associated disease data, comprising: A data processing module is used to collect disease status data information of obese patients and perform desensitization processing on the data information; A category labeling module is used to standardize the disease status data information and the preset nine TCM constitution identification feature data, and label the TCM constitution categories of obesity-associated diseases; The feature screening module is used to screen the feature data corresponding to the TCM constitution of obesity-associated diseases based on the IV value analysis method, obtain the feature data for model construction, and form a data set; A model building module, for using the data set to build a TCM constitution classification model integrating the combined disease data of obese patients based on a logistic regression algorithm; The category prediction module is used to predict the concomitant disease sample data of the patient to be tested using the TCM constitution classification model to obtain the obesity TCM constitution category.

[0010] The present invention further discloses a computer-readable storage medium, wherein the storage medium stores a computer program, and the computer program implements the above method when executed by a processor.

[0011] The present invention further discloses an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above method when executing the program.

[0012] The present invention further discloses a computer program product, comprising a computer program, which implements the above method when executed by a processor.

[0013] According to the technical solution of the present invention, by collecting the medical conditions and basic information of obese patients with accompanying diseases and performing desensitization processing, a basic data set covering multi-dimensional disease characteristics is constructed under the premise of ensuring data security; by performing type conversion, data filling and other standardization processing on the disease data and the nine constitution identification characteristics of traditional Chinese medicine and labeling a single constitution category, the data quality and the accuracy of model training are significantly improved. Furthermore, based on the IV value analysis method, feature data with high predictive value for constitution typing, such as diabetes and hypertension, are selected to form a data set, which effectively reduces feature redundancy, accelerates the convergence speed during model training and reduces time overhead. At the same time, the selected feature data are converted into probability values ​​through the Sigmoid function using the logistic regression algorithm, and the maximum likelihood estimation and gradient descent method are combined to construct a constitution typing model that integrates disease data. This not only achieves the probabilistic quantitative prediction of the constitution type of obese patients with accompanying diseases, but also reflects the degree of influence of different disease characteristics on constitution through linear combination coefficients, enhancing the interpretability and scientificity of the model; by performing the same preprocessing step on the test patient data and then inputting it into the model, the constitution category is determined according to the probability threshold, realizing an automated and accurate prediction process from data input to typing result output. It has achieved a deep integration of the combined disease data of obese patients with TCM constitution identification, avoiding the limitation of traditional constitution classification that does not consider disease specificity. It not only provides clinical practice with a personalized constitution identification tool that takes into account disease differences, but also can dynamically calibrate the model through continuous data supplementation, significantly improving the reliability of data analysis and classification accuracy, providing a scientific and quantitative basis for the formulation of TCM precise prevention and intervention plans for obesity and its complications, and has important clinical application value and economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] For purposes of illustration and not limitation, the present invention will now be described with reference to embodiments thereof and the accompanying drawings, in which: Figure 1 Schematic diagram of the process of the TCM constitution typing method based on obesity-associated disease data in an embodiment of the present invention; Figure 2 Schematic diagram of the composition structure of the TCM constitution classification system based on obesity-associated disease data in an embodiment of the present invention; Figure 3 Schematic diagram of the structure of an electronic device in an embodiment of the present invention. DETAILED DESCRIPTION

[0015] In order to enable those skilled in the art to better understand the solution of this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by ordinary technicians in this field without making creative work should fall within the scope of protection of this application.

[0016] It should be noted that, in the absence of conflict, the embodiments of the present application and the features thereof can be combined with each other. The embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0017] refer to Figure 1 This embodiment discloses a method for TCM constitution typing based on obesity-associated disease data, the method comprising the following steps: S100: Collecting disease status data information of obese patients and performing desensitization processing on the data information.

[0018] First, based on the clinical diagnosis and treatment system and electronic health records, data on obese patients' accompanying conditions such as diabetes, hypertension, coronary heart disease, hyperuricemia, gout, and related indicator data (including fasting blood glucose, blood pressure, blood uric acid levels, BMI index), and basic patient information data (including gender, current age, age of first symptom onset, weight, and occupation) are collected.

[0019] When desensitizing the original data, first remove the noise through data cleaning to convert qualitative variables into numerical variables, such as gender "male" is coded as 1, "female" is coded as 0, blood pressure ≥140 / 90mmHg is marked as 1, otherwise it is 0, and continuous indicators such as blood oxygen concentration and heart rate are divided into discrete values ​​according to preset intervals, such as blood oxygen [95-99%] is converted to 0, [90-94%] is converted to 1. Statistical methods are used to identify and correct outliers. In this embodiment, the Z-score method is used to eliminate outliers, missing values ​​are filled with the mean, and duplicate records are deleted. After the data cleaning is completed, a classification label is added to each test record based on the clinical diagnosis data. For example, BMI ≥ 28kg / m 2 Obese patients are labeled 1, and non-obese individuals are labeled 0, forming the initial dataset for model training. At the same time, sensitive personal information in the data, such as names and ID numbers, is irreversibly encrypted or anonymized to ensure compliance with data privacy regulations. This step preserves the integrity and unique value of obesity-related disease data while ensuring data security and compliance, laying the foundation for subsequent standardized processing and model building.

[0020] S200: Standardizing the disease condition data information and the preset nine TCM constitution identification feature data, and marking the TCM constitution categories of obesity-associated diseases.

[0021] The disease condition data information processed in step S100 and the preset nine constitution identification characteristic data of traditional Chinese medicine (such as the frequency of fatigue symptoms corresponding to qi deficiency constitution and the degree of thick and greasy tongue coating corresponding to phlegm-damp constitution) are uniformly processed in multi-source data format. In this embodiment, the nine constitutions of traditional Chinese medicine include: balanced constitution, qi deficiency constitution, yang deficiency constitution, yin deficiency constitution, phlegm-damp constitution, damp-heat constitution, blood stasis constitution, qi stagnation constitution, and special constitution. Specifically, binary data is discretized and mapped to numerical values. For example, the "Does one have high blood pressure?" field is mapped to "1 (yes) / 0 (no)" and the "Does one have a sedentary occupation?" field is mapped to "1 (yes) / 0 (no)." Continuous data is coded and assigned intervals according to clinical diagnosis and treatment guidelines. For example, blood oxygen concentration (95-99%, 90-94%, 85-89%, ≤84%) is converted to "0, 1, 2, 3," and heart rate (bpm) is converted to "0, 1, 2, 3, 4," achieving normalized representation of data with different dimensions. For null-valued fields within each health item, a mean-filling method is used. For example, missing values ​​are filled using the mean value of tongue coating thickness for people in the same disease group and physical constitution type to maintain data integrity. Furthermore, duplicate records and obvious contradictions are removed to further purify the dataset. Furthermore, a single TCM constitution category label is added to each obese patient sample, and the label set is limited to [balanced constitution, qi deficiency constitution, yang deficiency constitution, yin deficiency constitution, phlegm-damp constitution, damp-heat constitution, blood stasis constitution, qi stagnation constitution, special constitution]. For example, for patients with type 2 diabetes and BMI = 30 kg / m 2 Patients with predominant clinical symptoms of fatigue and spontaneous sweating were labeled as having a "Qi deficiency constitution." Patients with predominant symptoms of obesity, a greasy coating, and sticky stools were labeled as having a "phlegm-dampness constitution." This standardized process generated a formatted dataset with a unified structure and clear labels, providing high-quality input for subsequent feature screening and logistic regression modeling based on IV value analysis.

[0022] S300: Screening characteristic data corresponding to TCM constitutions associated with obesity-related diseases based on IV value analysis to obtain characteristic data for model construction and form a data set.

[0023] The IV value analysis method measures the impact of a feature on a target. Its basic idea is to compare and calculate the degree of correlation based on the ratio of black and white samples hit by the feature to the total black and white samples. Applying this IV value analysis to labeled stroke sample data yields filtered sample feature data, which helps reduce noise interference, accelerates convergence during model training, and reduces time overhead. For example, using health data from a specific target population, the IV value analysis method ultimately identified several features with a significant impact on obesity-related TCM constitutions as input variables for subsequent models, including diabetes, hypertension, coronary heart disease, hyperuricemia, and gout.

[0024] In this embodiment, first, the standardized disease condition data is associated with the nine constitution identification feature data of traditional Chinese medicine, and the marked obesity traditional Chinese medicine constitution category is used as the target variable to construct a mapping relationship between the feature and the constitution category. The IV value (information value) analysis method is used to evaluate the importance of each feature. Its core logic is: the IV value analysis method is used to measure the degree of influence of each feature on the obesity traditional Chinese medicine constitution category, and the correlation degree is calculated based on the ratio of black and white samples hit by the feature and the ratio of total black and white samples. Specifically including: Feature grouping: Continuous features (such as BMI and fasting blood glucose) are divided into reasonable intervals according to clinical significance (such as BMI ≥ 28 is the obesity interval), and discrete features (such as whether or not the patient has diabetes) are grouped by category; Correlation calculation: By comparing the distribution differences between target and non-target constitution samples in each feature group, the ability of the feature to distinguish constitution categories is quantified, and an IV value reflecting the predictive value of the feature is generated; Feature screening: retain features with higher IV values, eliminate redundant features that have little impact on constitution typing, and focus on screening out diabetes, hypertension, coronary heart disease, hyperuricemia, gout and other diseases and related indicator data as key features.

[0025] Through the above screening process, a data set containing features with high predictive value is formed, which effectively reduces the interference of noise information, reduces the time overhead of model training and accelerates convergence, providing streamlined and efficient feature input for subsequent logistic regression modeling.

[0026] S400: Based on a logistic regression algorithm, the data set is used to construct a TCM constitution classification model that integrates the combined disease data of obese patients.

[0027] The data set filtered by IV value analysis in step S300 is used as input, and the marked TCM constitution category is used as the target variable to construct a logistic regression model. The method for constructing the logistic regression model includes: Based on linear regression, the linear combination results of the characteristic data of the data set are mapped to the (0, 1) interval through the Sigmoid function, and the probability of predicting the obesity TCM constitution classification is calculated.

[0028] Specifically, in the logistic regression model, the dependent variable can only be a 0-1 binomial distribution, and the loss function of the model is the maximum likelihood estimation, that is, when given data, each parameter in the indicator variable is partially differentiated, and then the partial differential formula is set to 0, so as to find the estimated value of the parameter. Although the dependent variable can only be a 0-1 binomial distribution, since the loss function used is the maximum likelihood estimation, it is essentially calculating the probability that the dependent variable is 0 or the dependent variable is 1; therefore, the task of predicting a 0-1 classification variable can be converted into a model that predicts the probability of the variable being 1. In this embodiment, for example, the probability of the nth detection information and data having an obesity TCM constitution event is regarded as p n , then the probability of the TCM constitution event of obesity with label 1 is: Where Z = W T X=w0+w1x1+w2x2+…+w n x n , X=[x1,x2,...,x n ] is the feature index data set, W T is the optimal coefficient obtained by training the loss function. When Z approaches positive infinity, the probability p n Approaches 1, when Z approaches negative infinity, the probability p n Approaching 0. Similarly, the probability of the TCM constitution event of obesity with label 0 is: Therefore, the expression of the logistic regression model is:

[0029] The maximum likelihood estimation is used to construct the loss function. The logarithm of the likelihood function is taken and converted to minimize the negative log-likelihood function. The optimal coefficient vector is iteratively solved by the gradient descent method to converge the loss function, and finally the TCM constitution classification model is constructed. In the loss function, it is assumed that each characteristic index is independent of each other. According to the maximum likelihood function, the probability of the TCM constitution event occurring in the input data can be written as: Taking the logarithm on both sides of the above likelihood function, it is converted to: Taking the logarithm does not affect the monotonicity of the original function, but will significantly distinguish the differences between probabilities, and can better distinguish the categories of each sample. Multiplying the left and right sides by -1, converts it to minimize the negative log-likelihood function: Finally, the loss function is obtained by averaging the N samples to minimize the negative log-likelihood function:

[0030] Therefore, by taking the logarithm and then the negative number based on the 0-1 distribution, the loss function is matched with the distribution law of the 0-1 distribution. The minimum value of the function can be solved using the gradient descent method. Specifically, the partial derivative of the corresponding variable can be calculated based on the result after taking the logarithm and iterated step by step until convergence, and W can be obtained. T That is, the optimal coefficient obtained through loss function training. The final logistic regression risk probability prediction model for TCM constitution events in obese patients is:

[0031] S500: The concomitant disease sample data of the patient to be tested is predicted by the TCM constitution classification model to obtain the obesity TCM constitution category. Specifically, the concomitant disease sample data of the patient to be tested is standardized, and the sample data is converted into an input vector consistent with the feature data format in the data set; the input vector is input into the logistic regression model to calculate the probability value of the patient to be tested belonging to each TCM constitution category; according to a preset probability threshold, the constitution category with the highest probability value is determined as the obesity TCM constitution category of the patient to be tested.

[0032] refer to Figure 2 This embodiment further discloses a TCM constitution classification system based on obesity-associated disease data, including a data processing module 21, a category labeling module 22, a feature screening module 23, a model building module 24, and a category prediction module 25, which are further described below: The data processing module 21 is used to collect disease status data information of obese patients and desensitize the data information, including: collecting data on diabetes, hypertension, coronary heart disease, hyperuricemia, gout and other diseases and related indicators of obese patients, as well as basic information data such as the patient's gender, age, weight, occupation, etc.; cleaning the collected diagnostic test data, including converting qualitative variables into numerical variables, processing outliers and missing values, and deleting duplicate values; adding classification labels to the cleaned test data based on the diagnostic data to construct a data set that can be used by the model.

[0033] The category labeling module 22 is used to standardize the disease status data information and the preset nine TCM constitution identification feature data, and mark the TCM constitution category of obesity-associated diseases, including: type conversion, data filling and data deletion operations on the disease status data information and the nine TCM constitution identification feature data; marking the TCM constitution type of obesity-associated diseases as a single category, and the category labels correspond to the TCM constitution categories including: balanced constitution, qi deficiency constitution, yang deficiency constitution, yin deficiency constitution, phlegm-damp constitution, damp-heat constitution, blood stasis constitution, qi stagnation constitution, and special constitution.

[0034] The feature screening module 23 is used to screen the feature data corresponding to the TCM constitution of obesity-associated diseases based on the IV value analysis method, obtain feature data for model construction, and form a data set, including: measuring the degree of influence of each feature on the TCM constitution category of obesity through the IV value analysis method, and calculating the degree of correlation based on the ratio of black and white samples hit by the feature and the ratio of total black and white samples; screening out feature data that meets the prediction of the TCM constitution category of obesity, the feature data including the disease conditions and indicator data of diabetes, hypertension, coronary heart disease, hyperuricemia, and gout, and forming a data set for model construction based on the screened feature data.

[0035] The model construction module 24 is used to construct a TCM constitution classification model that integrates the combined disease data of obese patients using the data set based on a logistic regression algorithm, including: mapping the linear combination results of the characteristic data of the data set to the (0, 1) interval through the Sigmoid function based on linear regression, and calculating the probability of predicting the TCM constitution classification of obesity; using maximum likelihood estimation to construct a loss function, taking the logarithm of the likelihood function and converting it into a function that minimizes the negative log-likelihood, iteratively solving the optimal coefficient vector through the gradient descent method, so that the loss function converges, and finally constructing the TCM constitution classification model.

[0036] The category prediction module 25 is used to predict the concomitant disease sample data of the patient to be tested by using the TCM constitution classification model to obtain the obesity TCM constitution category, including: standardizing the concomitant disease sample data of the patient to be tested, converting the sample data into an input vector consistent with the feature data format in the data set; inputting the input vector into the logistic regression model to calculate the probability value of the patient to be tested belonging to each TCM constitution category; and according to a preset probability threshold, determining the constitution category with the highest probability value as the obesity TCM constitution category of the patient to be tested.

[0037] Figure 3 A schematic diagram of the physical structure of an electronic device provided in an embodiment of the present invention, such as Figure 3 As shown, the electronic device 50 includes: a processor 501 (processor), a memory 502 (memory) and a bus 503; The processor 501 and the memory 502 communicate with each other via the bus 503 ; the processor 501 is used to call program instructions in the memory 502 to execute the methods provided by the above-mentioned method implementation methods.

[0038] This embodiment provides a non-transitory computer-readable storage medium, which stores computer instructions. The computer instructions enable a computer to execute the methods provided by the above-mentioned method embodiments.

[0039] Those skilled in the art will understand that all or part of the steps for implementing the above-mentioned method implementation method can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above-mentioned method implementation method; and the aforementioned storage medium includes: ROM, RAM, disk or optical disk, etc. Various storage media that can store program codes.

[0040] The device embodiments described above are merely illustrative. Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of these modules may be selected based on actual needs to achieve the objectives of this embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0041] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, or of course, by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the existing technology can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of each embodiment or certain parts of the embodiment.

[0042] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may occur depending on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A method for TCM constitution typing based on obesity-associated disease data, characterized in that: The method comprises: Collecting disease status data and information of obese patients and performing desensitization processing on the data; Standardize the disease condition data and the preset nine TCM constitution identification feature data, and mark the TCM constitution categories of obesity-associated diseases; Based on the IV value analysis method, the characteristic data corresponding to the TCM constitution of obesity-associated diseases were screened to obtain the characteristic data for model construction and form a data set; Based on the logistic regression algorithm, the data set is used to construct a TCM constitution classification model that integrates the combined disease data of obese patients; the TCM constitution classification model is used to predict the concomitant disease sample data of the tested patients to obtain the TCM constitution category of obesity.

2. The TCM constitution typing method based on obesity-associated disease data according to claim 1, characterized in that: The collecting of disease status data information of obese patients and performing desensitization processing on the data information includes: Collect data on diabetes, hypertension, coronary heart disease, hyperuricemia, gout and other conditions and related indicators of obese patients, as well as basic information such as gender, age, weight, and occupation of the patients; Perform data cleaning on the collected diagnostic test data, including converting qualitative variables into numerical variables, processing outliers and missing values, and deleting duplicate values; Add classification labels to the cleaned detection data based on the diagnostic data to build a dataset that can be used by the model.

3. The TCM constitution typing method based on obesity-associated disease data according to claim 1, characterized in that: The disease condition data information and the preset TCM nine constitution identification feature data are standardized, and the categories of TCM constitutions associated with obesity-related diseases are marked, including: Type conversion, data filling and data deletion operations are performed on disease status data information and the nine constitution identification feature data of traditional Chinese medicine; the traditional Chinese medicine constitution types of obesity-associated diseases are marked as a single category, and the category labels correspond to the following traditional Chinese medicine constitution categories: balanced constitution, qi deficiency constitution, yang deficiency constitution, yin deficiency constitution, phlegm-damp constitution, damp-heat constitution, blood stasis constitution, qi stagnation constitution, and special constitution.

4. The TCM constitution typing method based on obesity-associated disease data according to claim 1, characterized in that: The IV value analysis method is used to screen the characteristic data corresponding to the TCM constitution of obesity-associated diseases to obtain characteristic data for model construction, which constitutes a data set including: The IV value analysis method was used to measure the influence of each feature on the TCM body type of obesity, and the correlation degree was calculated by comparing the ratio of black and white samples hit by the feature with the ratio of total black and white samples; Feature data that are consistent with the prediction of the TCM constitution category for obesity are screened out, including the disease conditions and indicator data of diabetes, hypertension, coronary heart disease, hyperuricemia, and gout, and a data set for model construction is formed based on the screened feature data.

5. The TCM constitution typing method based on obesity-associated disease data according to claim 1, characterized in that: The method of constructing a TCM constitution classification model integrating the combined disease data of obese patients using the data set based on a logistic regression algorithm comprises: mapping the linear combination result of the characteristic data of the data set to the interval (0, 1) through a Sigmoid function based on linear regression, and calculating the probability of predicting the TCM constitution classification of obesity; The maximum likelihood estimation is used to construct the loss function, the logarithm of the likelihood function is taken and converted into a minimized negative log-likelihood function, and the optimal coefficient vector is iteratively solved by the gradient descent method to make the loss function converge, and finally a TCM constitution classification model is constructed.

6. The TCM constitution typing method based on obesity-associated disease data according to claim 1, characterized in that: The TCM constitution classification model is used to predict the concomitant disease sample data of the patient to be tested, and the TCM constitution categories of obesity include: Standardizing the sample data of the patients with concomitant diseases to be tested, and converting the sample data into an input vector consistent with the format of the feature data in the dataset; The input vector is input into the logistic regression model to calculate the probability value of the patient to be tested belonging to each TCM constitution category; based on a preset probability threshold, the constitution category with the highest probability value is determined as the obese TCM constitution category of the patient to be tested.

7. A TCM constitution classification system based on obesity-associated disease data, characterized by: include: A data processing module is used to collect disease status data information of obese patients and perform desensitization processing on the data information; A category labeling module is used to standardize the disease status data information and the preset nine TCM constitution identification feature data, and label the TCM constitution categories of obesity-associated diseases; The feature screening module is used to screen the feature data corresponding to the TCM constitution of obesity-associated diseases based on the IV value analysis method, obtain the feature data for model construction, and form a data set; A model building module, for using the data set to build a TCM constitution classification model integrating the combined disease data of obese patients based on a logistic regression algorithm; The category prediction module is used to predict the concomitant disease sample data of the patient to be tested using the TCM constitution classification model to obtain the obesity TCM constitution category.

8. A computer-readable storage medium, characterized in that The storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the method according to any one of claims 1 to 6 is implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images