Analysis Method, System, Device and Storage Medium of Traditional Chinese Medicine Empirical Prescriptions

Abstract

The present invention relates to an analysis method, system, device and storage medium for traditional Chinese medicine empirical prescriptions, including: receiving first data, and obtaining a syndrome differentiation analysis result corresponding to the first data based on the theory of yin-yang and the five elements; mapping the syndrome differentiation analysis result into the twelve meridians to obtain a twelve-meridian vector corresponding to the first data, and obtaining second data; obtaining a twelve-meridian vector corresponding to the prescription based on the meridian tropism information of the prescription belonging to the twelve meridians, and obtaining a weighted value of the twelve-meridian vector corresponding to the prescription according to the basic composition of the prescription; performing weighted summation according to the corresponding twelve-meridian vector and the weighted value to obtain third data; inputting the second data and the third data into a trained first model to obtain a first result, and the first result is used to characterize the symptomatic degree of the prescription for the first data. The beneficial effect of the present invention is: improving the accuracy of the symptomatic degree of the prescription for the disease condition.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and particularly relates to an analysis method, system, device and storage medium for traditional Chinese medicine prescriptions. Background Art

[0002] Traditional Chinese medicine prescriptions are the accumulation of traditional Chinese medicine practice over thousands of years, with a vast amount of data, and are a brilliant treasure trove for studying traditional Chinese medicine. The so-called prescriptions include two aspects. One aspect is the record of the compatibility and usage of traditional Chinese medicine prescriptions, and the other aspect is the record of related symptoms. However, the recording methods of traditional Chinese medicine prescriptions are not standardized, so there are good and bad ones, and they must be analyzed and screened through scientific methods. Summary of the Invention

[0003] An object of the present invention is to solve at least one of the technical problems existing in the prior art, and provides an analysis method, system, device and storage medium for traditional Chinese medicine prescriptions, effectively improving the accuracy of prescriptions.

[0004] The technical solution of the present invention includes an analysis method, system, device and storage medium for traditional Chinese medicine prescriptions. The analysis method for traditional Chinese medicine prescriptions includes: receiving first data, and based on the theory of yin-yang and five elements, obtaining a syndrome differentiation analysis result corresponding to the first data, where the first data represents disease information;

[0005] Mapping the syndrome differentiation analysis result to the twelve meridians, obtaining a twelve-meridian vector corresponding to the first data to obtain second data; based on the meridian attribution information of the prescription belonging to the twelve meridians, obtaining a twelve-meridian vector corresponding to the prescription, and according to the basic composition of the prescription, obtaining a weighted value of the twelve-meridian vector corresponding to the prescription; performing weighted summation according to the corresponding twelve-meridian vector and the weighted value to obtain third data; inputting the second data and the third data into a trained first model to obtain a first result, where the first result is used to represent the symptomatic degree of the prescription for the first data.

[0006] According to the above-mentioned analysis method for traditional Chinese medicine prescriptions, the training of the first model includes: based on the meridian attribution information of the fourth data, obtaining first training data, where the fourth data includes at least a pair of prescription and disease information, and the symptomatic degree of the prescription in the fourth data for the disease information is greater than a first threshold; based on the meridian attribution information of the fifth data, obtaining second training data, where the fifth data includes at least a pair of prescription and disease information, and the symptomatic degree of the prescription in the fifth data for the disease information is less than a second threshold; training the first model based on the first training data and the second training data.

[0007] According to the analysis method of traditional Chinese medicine empirical formulae, obtaining the syndrome differentiation analysis result corresponding to the first data, including: obtaining the correlation probability value between the first data and the five elements by statistically analyzing the correlation probability information between diseases and the five elements, to obtain the sixth data; based on the sixth data, obtaining the information entropy of the five components of metal, wood, water, fire, and earth, where the information entropy of the component represents the degree of chaos of the component.

[0008] According to the analysis method of traditional Chinese medicine empirical formulae, obtaining the information entropy of the five components of metal, wood, water, fire, and earth, including: constructing a yin-yang orthogonal complex plane based on the cold, heat, deficiency, and excess of yin and yang; projecting the sixth data onto the real axis and the imaginary axis of the yin-yang orthogonal complex plane through frequency-time conversion to obtain the first result; based on the first result, obtaining the information entropy of the five components of metal, wood, water, fire, and earth in the time domain according to the entropy value calculation formula.

[0009] According to the analysis method of traditional Chinese medicine empirical formulae, obtaining the twelve-meridian vector corresponding to the first data, including: obtaining the components whose information entropy exceeds the set threshold; based on the components whose information entropy exceeds the set threshold, obtaining the meridian attribution information belonging to the twelve meridians to obtain the twelve-meridian vectors corresponding to at least one component; superimposing the twelve-meridian vectors corresponding to the at least one component to obtain the twelve-meridian vector corresponding to the first data.

[0010] According to the analysis method of traditional Chinese medicine empirical formulae, performing weighted summation according to the corresponding twelve-meridian vector and the weighting value to obtain the third data, including: obtaining the weight information of at least one medicinal material in the prescription according to the monarch, minister, assistant, and guide rules of the prescription; based on the meridian attribution information of the medicinal material belonging to the twelve meridians, obtaining the twelve-meridian vector corresponding to the medicinal material; based on the weight information of the medicinal material, performing weighted summation on the twelve-meridian vectors corresponding to at least one medicinal material to obtain the third data.

[0011] The technical solution of the present invention further includes an analysis system for traditional Chinese medicine prescriptions, comprising: a first module for receiving first data, obtaining a syndrome differentiation analysis result corresponding to the first data based on the theory of yin-yang and five elements, where the first data represents disease information; mapping the syndrome differentiation analysis result into the twelve meridians to obtain a twelve-meridian vector corresponding to the first data, and obtaining second data; a second module for obtaining a twelve-meridian vector corresponding to the prescription based on the meridian tropism information of the prescription belonging to the twelve meridians, obtaining a weighted value of the twelve-meridian vector corresponding to the prescription according to the basic composition of the prescription; performing weighted summation according to the corresponding twelve-meridian vector and the weighted value to obtain third data; a third module, the third module is connected to the first module, the third module is connected to the second module, and the third module is used to input the second data and the third data into a trained first model to obtain a first result, and the first result is used to represent the symptomatic degree of the prescription for the first data.

[0012] The technical solution of the present invention further includes an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and is characterized in that when the processor executes the computer program, the above-mentioned any method steps are implemented.

[0013] The technical solution of the present invention further includes a computer-readable storage medium, the computer-readable storage medium stores a computer program, and is characterized in that when the computer program is executed by a processor, the above-mentioned any method steps are implemented.

[0014] The beneficial effect of the present invention is that based on the syndrome differentiation theory, the disease information and prescription information are mapped in the twelve meridians, and the mapping relationship between the twelve-meridian vector corresponding to the disease information and the twelve-meridian vector corresponding to the prescription is learned by using a neural network, so as to improve the accuracy of the symptomatic degree of the prescription for the disease by reducing the complexity of the neural network model learning. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The present invention will be further described below in conjunction with the drawings and embodiments;

[0016] Figure 1 Shown is a block diagram of an embodiment of a system environment according to an embodiment of the present invention;

[0017] Figure 2 Shown is a flowchart according to an embodiment of the present invention;

[0018] Figure 3 Shown is a detailed flowchart according to an embodiment of the present invention;

[0019] Figure 4 Shown is a sub-flowchart according to an embodiment of the present invention;

[0020] Figure 5 Shown is another sub - flowchart according to an embodiment of the present invention;

[0021] Figure 6 Shown is a schematic diagram according to Embodiment 1 of the present invention;

[0022] Figure 7 Shown is the system structure diagram according to an embodiment of the present invention. Detailed Embodiments

[0023] This section will describe in detail the specific embodiments of the present invention. The preferred embodiments of the present invention are shown in the accompanying drawings. The role of the drawings is to supplement the description of the text part of the specification, enabling people to intuitively and vividly understand each technical feature and the overall technical solution of the present invention. However, it should not be construed as a limitation on the protection scope of the present invention.

[0024] In the description of the present invention, unless otherwise clearly defined, terms such as "set" should be understood in a broad sense. Those skilled in the art can reasonably determine the specific meaning of the above terms in the present invention in combination with the specific content of the technical solution.

[0025] As Figure 1 shown, the server may include one or more computing, storage, web, application, and / or other processing servers. The server may be located in one or more different data centers. The server may be instructed to collect disease and prescription information, train a machine learning model, calculate the degree of symptom matching between the prescription and the disease, etc. At least a part of the processing work executed by the server may include hosting and / or executing processing associated with an end - user request (e.g., providing the requested data / content to a computing system). Users access the services provided by one or more servers. For example, when a user inputs information about a certain disease case and the corresponding prescription information, at least a part of the processing work executed by the server may include hosting and / or executing processing associated with an end - user request and returning the degree of symptom matching between the prescription and the disease. Examples of users include personal computers, laptop computers, tablet computers, mobile devices, display devices, user input devices, and any other computing devices.

[0026] Although for simplicity of the diagram, only a limited number of component instances are shown, Figure 1 additional instances of any of the components shown in Figure 1Components not shown may also exist. The shown components communicate with each other via a network. Examples of networks include one or more of the following: direct or indirect physical communication connections, mobile communication networks, the Internet, intranets, local area networks, wide area networks, storage area networks, and any other form that connects two or more systems, components, or storage devices together.

[0027] As Figure 2 shown, a method for analyzing traditional Chinese medicine prescriptions includes the following steps:

[0028] S100, receive first data, and based on the theory of yin-yang and the five elements, obtain the syndrome differentiation analysis result corresponding to the first data, where the first data represents disease information;

[0029] S200, map the syndrome differentiation analysis result to the twelve meridians, obtain the twelve-meridian vector corresponding to the first data, and obtain second data;

[0030] S300, based on the meridian tropism information of the prescription belonging to the twelve meridians, obtain the twelve-meridian vector corresponding to the prescription, and according to the basic composition of the prescription, obtain the weighted value of the twelve-meridian vector corresponding to the prescription;

[0031] S400, perform weighted summation according to the corresponding twelve-meridian vector and the weighted value to obtain third data;

[0032] S500, input the second data and the third data into a trained first model to obtain a first result, where the first result is used to represent the symptomatic degree of the prescription for the first data.

[0033] In the method for analyzing traditional Chinese medicine prescriptions of the present invention, on the one hand, map the syndrome differentiation analysis result corresponding to the disease information to the twelve meridians to obtain the twelve-meridian vector corresponding to the disease information; on the other hand, based on the meridian tropism information of the prescription belonging to the twelve meridians and the basic composition of the prescription, obtain the twelve-meridian vector corresponding to the prescription; and obtain the symptomatic degree of the prescription for the disease information through the model, effectively improving the accuracy of the prescription.

[0034] The detailed implementation manners of the above steps are described in multiple embodiments below, as Figure 3 shown.

[0035] Before step S100, it further includes: collection and preprocessing of a training sample set, where the training sample set includes classical prescriptions and self-defined prescriptions.

[0036] In the process of doctors' syndrome differentiation and treatment, a series of prescription records will be generated, mainly including classical prescriptions and empirical prescriptions. Among them, classical prescriptions are prescriptions that conform to the theory of syndrome differentiation and treatment. Usually recorded in books, after a large number of practices, it has been verified that a certain prescription has a high degree of symptomatic treatment for a certain disease, and can be used by doctors under the guidance of the theory of syndrome differentiation and treatment; empirical prescriptions refer to prescriptions created by doctors in successive dynasties after the Han Dynasty. The recorded content generally includes: (1) the reasons for the generation of the prescription, such as the condition, symptoms, medical information collection, dialectical conclusion, etc.; (2) the analysis of the prescription, such as the basis of compatibility, dosage, and description of processing; (3) the record of curative effect and taboos for taking medicine, etc. Usually, the degree of symptomatic treatment of a certain prescription for a certain disease in an empirical prescription still needs to be confirmed. The present invention aims to construct an evaluation model for the degree of symptomatic treatment of prescriptions and diseases based on classical prescriptions or self-defined tagged prescriptions through the theory of syndrome differentiation and treatment of traditional Chinese medicine, so as to evaluate the degree of symptomatic treatment of prescriptions and diseases in empirical prescriptions.

[0037] (1) Collect a training sample set and label the samples, where the training sample set includes classical prescriptions and self-defined prescriptions.

[0038] It should be noted that the evaluation model for the degree of symptomatic treatment of prescriptions and diseases constructed in the present invention can be a regression model, outputting specific values of the degree of symptomatic treatment of prescriptions and diseases, such as 0% - 100%, or a classification model, outputting the levels of the degree of symptomatic treatment of prescriptions and diseases, such as three levels: low, medium, and high. For the convenience of description, the present invention is described in detail by way of examples, but is not limited to the content of the examples. For example, the evaluation model for the degree of symptomatic treatment of prescriptions and diseases constructed in the present invention is a classification model, in which the levels of the degree of symptomatic treatment of prescriptions and diseases are divided into two levels: low and high. For classical prescriptions, the degree of symptomatic treatment of prescriptions and diseases is high. In order to improve the generalization of the evaluation model for the degree of symptomatic treatment of prescriptions and diseases, the present invention self-defines prescriptions with a low degree of symptomatic treatment of prescriptions and diseases. For example, the disease information in classical prescriptions is recombined with the prescriptions, or the disease information is self-defined, and the prescriptions are self-defined against the theory of syndrome differentiation and treatment, so as to self-define prescriptions. The classical prescriptions and self-defined prescriptions are used as training data to train the model, as shown in Table 1.

[0039] Table 1. Training samples and corresponding labels

[0040]

[0041] Among them, the proportion of classical prescriptions and self-defined prescriptions in the training sample set can be freely selected. Generally, the number of samples of classical prescriptions is greater than or equal to the number of samples of self-defined prescriptions. In the present invention, it is exemplified that the proportion of classical prescriptions and self-defined prescriptions in the training sample set is 2:1.

[0042] (2) Convert classical prescriptions and custom prescriptions into word vectors in a unified format. Prescriptions are generally text data and have different formats. To help the server understand complex semantics, the present invention converts prescriptions into word vectors in a unified format through a word segmentation algorithm. Specifically, the present invention divides prescriptions according to the word segmentation algorithm and converts prescriptions into word vectors composed of at least one basic vector. Among them, for example, in the word segmentation of disease information, the basic disease vector is the smallest unit of disease description. For example, chest and hypochondrium, red eyes, and constipation can be used to construct a dictionary of disease descriptions based on the basic disease vector, and the disease information is segmented based on the dictionary; for example, in the word segmentation of prescription information, the basic medicinal material vector is the smallest unit of prescription description. For example, cassia seed, peony, and ginseng can be used to construct a dictionary of prescription descriptions based on the basic medicinal material vector, and the prescription information is segmented based on the dictionary. The process of word segmentation is a basic step in text processing. The present invention can use existing word segmentation algorithms to divide disease information and prescription information. For example, the Jieba word segmentation algorithm is used to segment the following classical prescription:

[0043] Disease information: "Pale complexion, low voice, shortness of breath, weakness, poor appetite and loose stools, pale tongue with white coating";

[0044] After word segmentation: [[pale complexion], [low voice], [shortness of breath], [poor appetite and loose stools], [pale tongue with white coating]]

[0045] Prescription information: "Atractylodes macrocephala, ginseng, poria cocos, and licorice";

[0046] After word segmentation: [[Atractylodes macrocephala], [ginseng], [poria cocos], [licorice]]

[0047] Label: high

[0048] S100, receive the first data, and based on the theory of yin-yang and five elements, obtain the syndrome differentiation analysis result corresponding to the first data. The first data represents disease information, specifically including:

[0049] S110, obtain the correlation probability value between the first data and the five elements by statistically analyzing the correlation probability information between diseases and the five elements, and obtain the sixth data, as Figure 4 shown;

[0050] S111. Construct a correlation probability table based on disease information and the five elements of yin and yang, with disease information as rows and the five elements of yin and yang as columns. Among them, the correlation probability table includes at least one probability value, and the disease information includes at least one basic disease vector. Let the first basic disease vector be any one of the basic disease vectors; the five elements of yin and yang include yin deficiency, yang deficiency, yin excess, yang excess, metal, wood, water, fire, and earth. The first object is any one of the five elements of yin and yang; the first basic disease vector and the first object determine the first probability value in the correlation probability table; the first probability value represents the probability that the disease in the first basic disease vector belongs to the first object. Based on the correlation table of disease information and the five elements of yin and yang, the probability value that a certain disease belongs to a specific object in the five elements of yin and yang can be queried. For example, if 1000 cases of disease information are collected, and based on the correlation table, it is found that there are 500 cases where red eyes belong to wood, then the probability that red eyes belong to wood is 0.5. For example, as shown in Table 1, it is the correlation probability table of disease information and the five elements of yin and yang. It should be noted that the present invention is not limited to using the correlation probability table to describe the correlation between disease information and the five elements of yin and yang, and can also be in any form such as graphs and vectors.

[0051] S112. Receive the first data. Based on the correlation probability table of disease information and the five elements of yin and yang, obtain the correlation probability value between the first data and the five elements, and obtain the sixth data. For example, the first data is: [[chest and hypochondrium distension and oppression], [bitter taste in the mouth and dryness in the throat], [red eyes]]. Based on Table 2, the sixth data is obtained, that is, vector B = [[a(1,1), a(1,2), a(1,3), a(1,4), a(1,5)], [a(2,1), a(2,2), a(2,3), a(2,4), a(2,5)], [a(3,1), a(3,2), a(3,3), a(3,4), a(3,5)]].

[0052] S120. Based on the sixth data, obtain the information entropy of the five components of metal, wood, water, fire, and earth. The information entropy of a component represents the degree of chaos of the component. Among them, the sixth data is regarded as discrete values in the frequency domain, as Figure 5 shown.

[0053] Table 2. Correlation table of disease information and the five elements of yin and yang

[0054]

[0055] S121. Construct a yin-yang orthogonal complex plane based on the cold, heat, deficiency, and excess of yin and yang;

[0056] Traditional Chinese medicine theory believes that the changes of yin and yang are continuous, and the changes acting on the human body represent the dynamic change trajectory of physical energy. Yin and yang are in opposition and unity. The present invention uses a continuous complex variable space to describe the laws of yin and yang. For example, as Figure 6As shown in the figure, taking yang as the real axis and yin as the imaginary axis, the positive semi-axis of the real axis represents yang excess / reality, the negative semi-axis represents yang decline / vacuity, the positive semi-axis of the imaginary axis represents yin excess / reality, and the negative semi-axis represents yin decline / vacuity, to construct a yin-yang orthogonal complex plane.

[0057] S122, project the sixth data onto the real axis and the imaginary axis of the yin-yang orthogonal complex plane through frequency-time conversion to obtain a first result; among them, frequency-time conversion can adopt methods such as inverse Fourier transform and inverse wavelet transform to convert the frequency-domain signal into a time-domain signal. For example, the discrete numerical value A in the frequency domain 1 , denoted as the function A′ 1 (w), and obtain the function a in the time domain through inverse Fourier transform 2 (t), as shown in formula (1);

[0058]

[0059] The discrete numerical value A in the frequency domain 1 , after frequency-time conversion, is projected onto the real axis and the imaginary axis of the yin-yang orthogonal complex plane to obtain the first result a 2 (t).

[0060] In one embodiment, the sixth data is the discrete numerical value B in the frequency domain, where the vector B = [b 1 , b 2 , b 3 , b 4 , b 5 , where the vectors b 1 , b 2 , b 3 , b 4 , b 5 are respectively the discrete numerical values of metal, wood, water, fire, and earth in the frequency domain, as shown in formula (1). The vector b k , through inverse Fourier transform, obtains the function b′ in the time domain k (t), where k = 1,..., 5; b k , after frequency-time conversion, is projected onto the real axis and the imaginary axis of the yin-yang orthogonal complex plane to obtain the first result b′ k (t).

[0061] S123, based on the first result, according to the entropy value calculation formula, obtain the information entropy of the five components of metal, wood, water, fire, and earth in the time domain. In the existing technology, to calculate the information entropy of a continuous signal, usually methods such as approximate entropy based on conditional probability calculation, sample entropy, fuzzy entropy, power spectrum entropy based on power spectrum calculation, singular spectrum entropy calculated based on singular value decomposition, and energy entropy calculated based on the energy of signal decomposition components can be used. The present invention can adopt any one of the calculation methods. For example, adopt approximate entropy calculated based on conditional probability to calculate the information entropy of b′ k (t) to obtain the information entropy of the kth component The information entropy of the five components of gold, wood, water, fire, and earth in the time domain can be obtained, as shown in Table 3.

[0062] Table 3. Information Entropy of the Five Elements

[0063]

[0064] For S200, map the result of syndrome differentiation analysis to the twelve meridians, obtain the twelve-meridian vector corresponding to the first data, and obtain the second data, specifically including:

[0065] For S210, obtain the components whose information entropy exceeds the set threshold; among them, the threshold can be set artificially or use the empirical value 0.5. For example, set the threshold to 0.5, and calculate the information entropy corresponding to each component based on step S123, as shown in Table 4, and the entropy value vector of the five elements is [0.4, 0.8, 0.5, 0.6, 0.5].

[0066] Table 4. Component Information Entropy of the Five Elements

[0067] Gold Wood Water Fire Earth 0.4 0.8 0.5 0.6 0.5

[0068] Among them, the entropy value of the five-element component characterizes the degree of pathological chaos. The larger the entropy value, the higher the degree of pathological chaos. As can be seen from Table 4, the degrees of pathological chaos of wood and fire are relatively large.

[0069] For S220, based on the components whose information entropy exceeds the set threshold, obtain the meridian attribution information belonging to the twelve meridians, and obtain the twelve-meridian vector corresponding to at least one component;

[0070] The present invention does not directly use a neural network to learn the prescription information corresponding to the disease, nor learn the correlation between the disease and the prescription text. Instead, based on the syndrome differentiation theory, the disease information and the prescription information are mapped in the twelve meridians, and a neural network is used to learn the mapping relationship between the twelve-meridian vector corresponding to the disease information and the twelve-meridian vector corresponding to the prescription, reducing the complexity of the neural network model learning and further improving the accuracy of the symptomatic degree of the prescription for the disease.

[0071] In traditional Chinese medicine theory, the Yellow Emperor's Inner Canon defines that there are 12 human meridians, called the twelve meridians. Therefore, the meridian attribution output vector is 12-dimensional, and each dimension represents a meridian. In the traditional Chinese medicine syndrome differentiation theory, each component of the five elements has the meridian attribution information of the twelve meridians. For example, as can be seen from Table 4, the degrees of pathological chaos of wood and fire are relatively large. The meridian attribution of the component wood is the liver, and the twelve-meridian vector corresponding to the component wood is s 2 =[r 1 ,....,r 12 , in the twelve meridians s 2 , the eigenvalue corresponding to the liver is 1, and the other eigenvalues are 0. The meridian attribution of the component fire is the heart, and the twelve-meridian vector corresponding to the component fire is s4 = [r' 1 ,...., r 1 ' 2 , among the twelve regular meridians s 4 In, the eigenvalue corresponding to the heart is 1, and the rest of the eigenvalues are 0.

[0072] S230. Superimpose the twelve - meridian vectors corresponding to at least one component to obtain the twelve - meridian vector corresponding to the first data.

[0073] For example, as can be seen from Table 4, the morbid chaos degrees of wood and fire are relatively large. The twelve - meridian vector corresponding to the component wood is s 2 = [r 1 ,...., r 12 ; the twelve - meridian vector corresponding to the component fire is s 4 = [r' 1 ,...., r' 12 . Superimpose the twelve - meridian vectors corresponding to the component wood and the component fire to obtain the twelve - meridian vector S = [r'' 1 ,...., r'' 12 , where the eigenvalue corresponding to the liver in the twelve - meridian S is 1, the eigenvalue corresponding to the heart is 1, and the rest of the eigenvalues are 0.

[0074] S300. Based on the meridian - tropism information of the prescription belonging to the twelve regular meridians, obtain the twelve - meridian vector corresponding to the prescription, and according to the basic composition of the prescription, obtain the weighted value of the twelve - meridian vector corresponding to the prescription, specifically including:

[0075] S310. According to the monarch - minister - assistant - guide rule of the prescription, obtain the weight information of at least one medicinal material in the prescription;

[0076] In the theory of traditional Chinese medicine syndrome differentiation, generally, the monarch - minister - assistant - guide is used to divide the importance of medicinal materials into 4 levels, with the monarch drug and the minister drug as the main consideration objects; the importance of the assistant drug and the guide drug is relatively much lower. Based on the monarch - minister - assistant - guide rule, this invention divides the weights of medicinal materials. For example, the monarch drug and the minister drug are the main consideration objects, and the proportion of such medicinal materials is relatively large; the importance of the assistant drug and the guide drug is relatively much lower, and the proportion of such medicinal materials is relatively small. For example, when the medicinal materials are the monarch, minister, assistant, and guide respectively, the proportion information is shown in Table 5.

[0077] Table 5. Division of the proportion of medicinal materials

[0078] King Minister Assistant Guide 0.5 0.3 0.1 0.1

[0079] For example, the medicine vector is: [[Atractylodes macrocephala], [Ginseng], [Poria cocos], [Licorice root]], where Atractylodes macrocephala is the sovereign drug. As can be seen from the table, the weight of Atractylodes macrocephala is 0.5, Ginseng is the ministerial drug with a weight of 0.3, Poria cocos is the adjuvant drug, and Licorice root is the guiding drug. The weights of Atractylodes macrocephala and Licorice root are 0.1 respectively.

[0080] S320. Based on the meridian tropism information of the medicinal materials belonging to the twelve meridians, obtain the twelve-meridian vectors corresponding to the medicinal materials.

[0081] In traditional Chinese medicine syndrome differentiation theory, each medicinal material in the prescription has meridian tropism information of the twelve meridians. For example, the medicine vector is: [[Atractylodes macrocephala], [Ginseng], [Poria cocos], [Licorice root]], and its meridian tropism is the heart, lungs, spleen, stomach, and kidneys. For example, the meridian tropism of Atractylodes macrocephala is the spleen and stomach, and the twelve-meridian vector corresponding to Atractylodes macrocephala is In the twelve meridians t 1 Among them, the characteristic values corresponding to the spleen and stomach are 1, and the rest of the characteristic values are 0; in addition, the meridian tropism of Ginseng is the spleen, lungs, and heart, and the twelve-meridian vector corresponding to Ginseng is In the twelve meridians t 1 Among them, the characteristic values corresponding to the spleen, lungs, and heart are 1, and the rest of the characteristic values are 0; the meridian tropism of Poria cocos is the heart, spleen, and kidneys, and the twelve-meridian vector corresponding to Poria cocos is In the twelve meridians t 1 Among them, the characteristic values corresponding to the heart, spleen, and kidneys are 1, and the rest of the characteristic values are 0; the meridian tropism of Licorice root is the heart, lungs, spleen, and stomach.

[0082] S400. Perform weighted summation according to the corresponding twelve-meridian vectors and weighted values to obtain the third data.

[0083] S410. Based on the weight information of the medicinal materials, perform weighted summation on the twelve-meridian vectors corresponding to at least one medicinal material to obtain the third data. Among them, performing weighted summation on the twelve-meridian vectors corresponding to at least one medicinal material is shown in formula (2).

[0084]

[0085] Among them, T represents performing weighted summation on the twelve-meridian vectors corresponding to at least one medicinal material, i represents the i-th medicinal material, M represents the total number of medicinal materials in the prescription, v i represents the weight of the i-th medicinal material, and t i represents the vector of the twelve meridians corresponding to the i-th medicinal material.

[0086] S500. Input the second data and the third data into the trained first model to obtain the first result. The first result is used to characterize the symptomatic degree of the prescription for the first data, specifically including:

[0087] S510. Based on the meridian tropism information of the fourth data, obtain the first training data. The fourth data includes at least a pair of prescription and disease information, and the symptomatic degree of the prescription in the fourth data for the disease information is greater than the first threshold.

[0088] In one embodiment, the evaluation model for constructing the symptomatic degree of the prescription and the disease in the present invention can be a regression model, which outputs the specific value of the symptomatic degree of the prescription for the disease, such as 0% to 100%. The first threshold can be set as the symptomatic degree of the prescription for the disease being 95%. The evaluation model for constructing the symptomatic degree of the prescription and the disease in the present invention can also be a classification model, which outputs the level of the symptomatic degree of the prescription and the disease. For example, there are two levels: low and high, represented by numerical values 0 and 1. The first threshold can be set as the symptomatic degree of the prescription for the disease being 0.8.

[0089] The fourth data can be classical prescriptions. The symptomatic degree of the prescriptions in the classical prescriptions for the diseases is all 1. By mapping the disease information and the prescriptions in the classical prescriptions to the twelve meridians respectively, vectors S and T are obtained, which are the first training data.

[0090] S520. Based on the meridian tropism information of the fifth data, obtain the second training data. The fifth data includes at least a pair of prescription and disease information, and the symptomatic degree of the prescription in the fifth data for the disease information is less than the second threshold.

[0091] In one embodiment, the evaluation model for constructing the symptomatic degree of the prescription and the disease in the present invention can be a regression model, which outputs the specific value of the symptomatic degree of the prescription for the disease, such as 0% to 100%. The second threshold can be set as the symptomatic degree of the prescription for the disease being 5%. The evaluation model for constructing the symptomatic degree of the prescription and the disease in the present invention can also be a classification model, which outputs the level of the symptomatic degree of the prescription and the disease. For example, there are two levels: low and high, represented by numerical values 0 and 1. The second threshold can be set as the symptomatic degree of the prescription for the disease being 0.2.

[0092] The fifth data can be customized prescriptions. The symptomatic degree of the prescriptions in the customized prescriptions for the diseases is all 0. By mapping the disease information and the prescriptions in the customized prescriptions to the twelve meridians respectively, vectors S' and T' are obtained, which are the second training data.

[0093] S530. Based on the first training data and the second training data, train the first model. Among them, the first model can adopt any neural network algorithm to construct a regression or classification model for the symptomatic degree of the prescription and the disease, which will not be listed one by one here.

[0094] As Figure 7 shown, an analysis system for traditional Chinese medicine empirical prescriptions includes:

[0095] The first module is used to receive the first data, obtain the syndrome differentiation analysis result corresponding to the first data based on the theory of yin-yang and five elements, where the first data represents disease information; map the syndrome differentiation analysis result into the twelve meridians, obtain the twelve-meridian vector corresponding to the first data, and obtain the second data;

[0096] The second module is used to obtain the twelve-meridian vector corresponding to the prescription based on the meridian tropism information of the prescription belonging to the twelve meridians, and obtain the weighted value of the twelve-meridian vector corresponding to the prescription according to the basic composition of the prescription; perform weighted summation according to the corresponding twelve-meridian vector and the weighted value to obtain the third data;

[0097] The third module, which is connected to the first module and the second module, is used to input the second data and the third data into the trained first model to obtain the first result, and the first result is used to represent the symptomatic degree of the prescription for the first data.

[0098] It should be recognized that the method steps in the embodiments of the present invention can be implemented or carried out by computer hardware, a combination of hardware and software, or computer instructions stored in a non-transitory computer-readable memory. The method can use standard programming techniques. Each program can be implemented in a high-level procedural or object-oriented programming language to communicate with the computer system. However, if necessary, the program can be implemented in assembly or machine language. In any case, the language can be a compiled or interpreted language. In addition, for this purpose, the program can run on a dedicated integrated circuit programmed for this purpose.

[0099] In addition, the operations of the processes described herein can be performed in any suitable order, unless otherwise indicated herein or otherwise clearly contradicted by the context. The processes described herein (or variations and / or combinations thereof) can be executed under the control of one or more computer systems configured with executable instructions, and can be implemented as code (e.g., executable instructions, one or more computer programs, or one or more applications) executed commonly on one or more processors, by hardware, or a combination thereof. The computer program includes a plurality of instructions executable by one or more processors.

[0100] Further, the method can be implemented in any type of computing platform operatively connected to a suitable one, including but not limited to personal computers, minicomputers, mainframes, workstations, network or distributed computing environments, separate or integrated computer platforms, or communicating with charged particle tools or other imaging devices, etc. Aspects of the present invention can be implemented in machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into the computing platform, such as a hard disk, optical read and / or write storage medium, RAM, ROM, etc., such that it can be read by a programmable computer and, when the storage medium or device is read by the computer, can be used to configure and operate the computer to perform the processes described herein. In addition, the machine-readable code, or portions thereof, can be transmitted via a wired or wireless network. When such media includes instructions or programs that implement the above-described steps in conjunction with a microprocessor or other data processor, the invention described herein includes these and other different types of non-transitory computer-readable storage media. When programmed according to the methods and techniques of the present invention, the present invention also includes the computer itself.

[0101] A computer program can be applied to input data to perform the functions described herein, thereby transforming the input data to generate output data stored in non-volatile memory. The output information can also be applied to one or more output devices such as a display. In a preferred embodiment of the present invention, the transformed data represents physical and tangible objects, including a specific visual depiction of the physical and tangible objects produced on a display.

[0102] The embodiments of the present invention have been described in detail above in conjunction with the accompanying drawings. However, the present invention is not limited to the above embodiments, and various changes can be made without departing from the spirit of the present invention within the scope of knowledge possessed by those of ordinary skill in the art.

Claims

1. A method for analyzing traditional Chinese medicine empirical prescriptions, characterized in that, it includes: Receiving first data, and based on the theory of yin-yang and the five elements, obtaining a syndrome differentiation analysis result corresponding to the first data, where the first data represents disease information; Mapping the syndrome differentiation analysis result to the twelve meridians, obtaining a twelve-meridian vector corresponding to the first data, and obtaining second data; Based on the meridian tropism information of the prescription belonging to the twelve meridians, obtaining a twelve-meridian vector corresponding to the prescription, and according to the basic composition of the prescription, obtaining a weighted value of the twelve-meridian vector corresponding to the prescription; Performing weighted summation according to the corresponding twelve-meridian vector and the weighted value to obtain third data; Inputting the second data and the third data into a trained first model to obtain a first result, where the first result is used to represent the symptomatic degree of the prescription for the first data; The training of the first model includes: Based on the meridian tropism information of the fourth data, obtaining first training data, where the fourth data includes at least a pair of prescription and disease information, and the symptomatic degree of the prescription in the fourth data for the disease information is greater than a first threshold; Based on the meridian tropism information of the fifth data, obtaining second training data, where the fifth data includes at least a pair of prescription and disease information, and the symptomatic degree of the prescription in the fifth data for the disease information is less than a second threshold; Training the first model based on the first training data and the second training data; The obtaining of the syndrome differentiation analysis result corresponding to the first data includes: By statistically calculating the probability information related to the five elements of the disease, obtaining a correlation probability value between the first data and the five elements, and obtaining sixth data; Based on the sixth data, obtaining the information entropy of the five components of metal, wood, water, fire, and earth, where the information entropy of the component represents the degree of chaos of the component; The obtaining of the information entropy of the five components of metal, wood, water, fire, and earth includes: Based on the cold, heat, deficiency, and excess of yin and yang, constructing a yin-yang orthogonal complex plane, where yang is the real axis, yin is the imaginary axis, the positive half-axis of the real axis is yang excess / strength, the negative half-axis is yang decline / deficiency, the positive half-axis of the imaginary axis is yin excess / strength, and the negative half-axis is yin decline / deficiency; Projecting the sixth data onto the real axis and the imaginary axis of the yin-yang orthogonal complex plane through frequency-time conversion to obtain a first result; Based on the first result, according to the entropy value calculation formula, obtaining the information entropy of the five components of metal, wood, water, fire, and earth in the time domain; The obtaining of the twelve-meridian vector corresponding to the first data includes: Obtaining the components whose information entropy exceeds a set threshold; Based on the components whose information entropy exceeds the set threshold, obtaining the meridian tropism information belonging to the twelve meridians, and obtaining the twelve-meridian vectors corresponding to at least one component; Superposing the twelve-meridian vectors corresponding to the at least one component to obtain the twelve-meridian vector corresponding to the first data.

2. The method for analyzing traditional Chinese medicine empirical prescriptions according to claim 1, characterized in that, The performing of weighted summation according to the corresponding twelve-meridian vector and the weighted value to obtain third data includes: According to the monarch, minister, assistant, and envoy rules of the prescription, obtaining the weight information of at least one medicinal material in the prescription; Based on the meridian tropism information of the medicinal materials belonging to the twelve meridians, obtain the twelve-meridian vectors corresponding to the medicinal materials; Based on the weight information of the medicinal materials, perform weighted summation on the twelve-meridian vectors corresponding to at least one medicinal material to obtain the third data.

3. An analysis system for traditional Chinese medicine prescriptions Characterized in that It includes: The first module is used to receive the first data, and based on the theory of yin-yang and five elements, obtain the syndrome differentiation analysis result corresponding to the first data, and the first data represents disease information; Map the syndrome differentiation analysis result to the twelve meridians, obtain the twelve-meridian vector corresponding to the first data, and obtain the second data; The second module is used to obtain the twelve-meridian vector corresponding to the prescription based on the meridian tropism information of the prescription belonging to the twelve meridians, obtain the weighted value of the twelve-meridian vector corresponding to the prescription according to the basic composition of the prescription; perform weighted summation according to the corresponding twelve-meridian vector and the weighted value to obtain the third data; The third module, the third module is connected to the first module, the third module is connected to the second module, and the third module is used to input the second data and the third data into the trained first model to obtain the first result, and the first result is used to characterize the symptomatic degree of the prescription for the first data; The training of the first model includes: Based on the meridian tropism information of the fourth data, obtain the first training data, and the fourth data includes at least a pair of prescription and disease information, and the symptomatic degree of the prescription in the fourth data for the disease information is greater than the first threshold; Based on the meridian tropism information of the fifth data, obtain the second training data, and the fifth data includes at least a pair of prescription and disease information, and the symptomatic degree of the prescription in the fifth data for the disease information is less than the second threshold; Based on the first training data and the second training data, train the first model; The obtaining of the syndrome differentiation analysis result corresponding to the first data includes: By statistically analyzing the correlation probability information between the disease and the five elements, obtain the correlation probability value between the first data and the five elements, and obtain the sixth data; Based on the sixth data, obtain the information entropy of the five components of metal, wood, water, fire, and earth, and the information entropy of the component characterizes the degree of chaos of the component; The obtaining of the information entropy of the five components of metal, wood, water, fire, and earth includes: Based on the cold, heat, deficiency, and excess of yin and yang, construct a yin-yang orthogonal complex plane, where the yang is the real axis, the yin is the imaginary axis, the positive half-axis of the real axis is yang excess / reality, the negative half-axis is yang decline / deficiency, the positive half-axis of the imaginary axis is yin excess / reality, and the negative half-axis is yin decline / deficiency; Project the sixth data onto the real axis and the imaginary axis of the yin-yang orthogonal complex plane through frequency-time conversion to obtain the first result; Based on the first result, according to the entropy value calculation formula, obtain the information entropy of the five components of metal, wood, water, fire, and earth in the time domain; The obtaining of the twelve-meridian vector corresponding to the first data includes: Obtain the components whose information entropy exceeds the set threshold; Based on the components whose information entropy exceeds the set threshold, obtain the meridian tropism information belonging to the twelve meridians, and obtain the twelve-meridian vectors corresponding to at least one component; Superimpose the twelve meridian vectors corresponding to the at least one component to obtain the twelve meridian vectors corresponding to the first data.

4. An electronic device, the device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein, when the processor executes the computer program, the method steps described in any one of claims 1-2 are implemented.

5. A computer-readable storage medium, the computer-readable storage medium stores a computer program, wherein, when the computer program is executed by a processor, the method steps described in any one of claims 1-2 are implemented.

Images