A Traditional Chinese Medicine Prescription Generation System Based on Data Analysis

Abstract

This invention relates to the field of data analysis technology, and more particularly to a data analysis-based system for generating traditional Chinese medicine (TCM) prescriptions. An information acquisition module collects feature description sets containing sensitive sub-features, general sub-features, and quantifiable sub-features. A database module stores the feature description sets, quantifiable sub-feature data, state labels, and final combinations for each feature. A data integration module performs basic matching and preliminary label classification on the feature description sets, analyzes quantifiable sub-features to determine feature similarity types, and integrates the final combinations into a state storage type. An output module generates an initial combination based on the feature description set of the analysis target and outputs the final combination. A model analysis module adjusts the initial combination using a pharmacological database and sensitive sub-features to optimize the safety and effectiveness of the prescription. This system enables intelligent and personalized generation of TCM prescriptions, improving the accuracy and rationality of TCM diagnosis and treatment.

Description

Technical Field

[0001] This invention relates to the field of data analysis technology, and in particular to a traditional Chinese medicine prescription generation system based on data analysis. Background Technology

[0002] In existing technologies, prescription generation is typically based on a retrieval-enhanced Seq2Seq learning model called PreGenerator. This model comprises three modules: symptom-prescription retrieval, herb-herb retrieval, and prescription generation. It first uses patient symptom characteristics combined with syndrome differentiation as a query to retrieve the most relevant prescriptions from a classical prescription retrieval pool, generating prescription-level feature templates. Simultaneously, the herb-herb retrieval module analyzes the relevance of herbs in the retrieved prescriptions, generating herb-level templates. Finally, a prescription decoder combines symptom features, prescription-level templates, and herb-level template feature representations to generate a rigorous prescription. This approach not only considers symptoms and medical history but also incorporates traditional Chinese medicine prescription compatibility rules, resulting in more professional and persuasive prescriptions.

[0003] While existing TCM prescription generation technologies have made progress in terms of intelligence and personalization, many problems still exist. On the one hand, they rely on large-scale, high-quality data. When data is insufficient or of poor quality, the generated prescriptions may lack accuracy and be difficult to effectively treat new symptoms or rare diseases, limiting their generalization ability. On the other hand, the generated prescriptions sometimes contain inappropriate drug combinations and other irrationalities, posing potential safety risks. Furthermore, they are not precise enough in terms of personalized adjustments, making it difficult to fully meet the individual needs of different patients. Summary of the Invention

[0004] Therefore, this invention provides a data analysis-based TCM prescription generation system to overcome the problem of insufficient precision in personalized adjustments such as symptom-based prescriptions in existing technologies.

[0005] To achieve the above objectives, the preferred technical solution for a data analysis-based TCM prescription generation system is as follows:

[0006] This invention provides a data analysis-based traditional Chinese medicine prescription generation system, comprising:

[0007] The information acquisition module is used to acquire a feature description set of the analysis target, wherein the feature description set includes sensitive sub-features, general sub-features and quantifiable sub-features;

[0008] The database module stores the complete feature description set corresponding to each feature individual, the specific data of the quantifiable sub-features in each feature description set, and the status label corresponding to each feature description set; and records the final combination corresponding to the complete feature description set of each feature individual, wherein the final combination includes the final component and the corresponding component content;

[0009] The data integration module, which is connected to the database module, is used to perform basic matching on all the feature description sets to classify all feature individuals into preliminary labels, and to analyze the quantifiable sub-features in each preliminary label category to determine the feature similarity type, and to combine the state labels to integrate the final combination of the corresponding feature similarity types into a state storage type.

[0010] The output module is connected to the information acquisition module and the data integration module respectively, and is used to analyze the feature description set of the analysis target to obtain an initial combination, and output the final combination based on the analysis results of the initial combination;

[0011] The model analysis module, which is connected to the data integration module and the output module respectively, is used to adjust the initial combination based on the pharmacological database and the sensitivity sub-features of the analysis target.

[0012] As a preferred technical solution for a data analysis-based TCM prescription generation system, the database module stores several feature individuals and their complete feature description sets, wherein the number of feature description sets is greater than or equal to the total number of feature individuals.

[0013] As a preferred technical solution for a data analysis-based TCM prescription generation system, the data integration module performs preliminary labeling and classification of all feature individuals based on the general sub-features and the quantifiable sub-features.

[0014] If any of the quantifiable sub-features exceeds the corresponding threshold, and / or the matching result of the general sub-features with the chronic feature database is lower than the corresponding threshold, then the data integration module determines that the category of the preliminary label classification of the feature individual is the acute prescription category.

[0015] If there are quantifiable sub-features that do not exceed the corresponding threshold and the matching result of the general sub-features and the chronic feature database exceeds the corresponding threshold, then the data integration module determines that the category of the preliminary label classification of the feature individual is the chronic prescription category.

[0016] As a preferred technical solution for a data analysis-based TCM prescription generation system, the data integration module analyzes the quantifiable sub-features in each preliminary label category to determine several feature similarity types.

[0017] All quantizable features are divided into several feature similarity types using the corresponding thresholds of each quantizable sub-feature;

[0018] Among them, different feature similarity types have some of the same quantifiable sub-features.

[0019] As a preferred technical solution for a data analysis-based TCM prescription generation system, the output module generates an initial combination based on the feature description set of the analysis target:

[0020] If the analysis target is classified as an acute prescription category, the output module matches the similarity between its quantifiable sub-features and each combination in the acute state storage type, selects several final combinations with the highest similarity as candidate combinations, and calculates the component content of each candidate combination by weighting according to the degree of abnormality of the quantifiable sub-features.

[0021] If the analysis target is classified as a chronic prescription category, the output module matches the correlation between its general sub-features and the core component combination in the chronic state storage type, and adjusts the basic component allocation ratio according to the matching results of the chronic feature database.

[0022] As a preferred technical solution for a TCM prescription generation system based on data analysis, the degree of anomaly of the quantifiable sub-feature is determined by the interval difference between the quantifiable sub-feature of the current analysis target and the corresponding feature similarity type in the state storage type.

[0023] The correlation degree is determined based on the number of general sub-features and the degree of consistency between the combinations in the chronic state storage type.

[0024] As a preferred technical solution for a data analysis-based TCM prescription generation system, the model analysis module screens each component in the initial combination according to the sensitive sub-features to obtain screened components, and determines alternative components according to the pharmacological database.

[0025] The model analysis module obtains a second combination based on the screened components and the alternative components.

[0026] As a preferred technical solution for a data analysis-based TCM prescription generation system, the model analysis module determines the rationality of the second combination based on the pharmacological database, and determines the final combination based on the rationality judgment result;

[0027] If the rationality judgment result of the second combination is normal, then the components of the final combination determined by the model analysis module are the same as those of the second combination;

[0028] If the rationality judgment result of the second combination is abnormal, the model analysis module determines to repeat the process of determining alternative components and corresponding rationality based on the pharmacological database.

[0029] The beneficial effects of this invention are as follows:

[0030] This invention provides a comprehensive understanding of a patient's specific condition by acquiring a feature description set of the analysis target, including sensitive sub-features, general sub-features, and quantifiable sub-features. The database module stores a rich collection of individual features and their corresponding feature description sets and status labels, providing a solid data foundation for subsequent analysis. The data integration module performs basic matching and analysis on the feature description sets, and preliminary label classification improves the accuracy and efficiency of the analysis. The output module generates initial combinations based on the analysis results and further outputs final combinations, enabling rapid prescription recommendations. The model analysis module adjusts the initial combinations based on a pharmacological database and sensitive sub-features, ensuring the safety and effectiveness of the prescriptions. This invention not only improves the scientific rigor and rationality of TCM prescription generation but also effectively enhances the personalization and precision of TCM prescriptions, providing patients with safer and more effective treatment options, and also strongly supports the modernization and intelligent development of TCM.

[0031] In particular, the model analysis module screens components in the initial combination based on sensitive sub-features and identifies alternative components using a pharmacological database, thus obtaining a second combination. This process reflects a precise consideration of individual patient differences and avoids the negative impact of contraindications such as allergens on treatment efficacy. Simultaneously, by judging the rationality of the second combination, it ensures that the final combination conforms to pharmacological principles and maximizes therapeutic effects. If the rationality judgment result is normal, the final combination is directly determined to be the same as the second combination, ensuring the stability and reliability of the treatment plan. If an abnormality occurs, the process of identifying alternative components and judging rationality is repeated to continuously optimize the combination until a reasonable final combination is obtained. This mechanism effectively improves the scientific nature and safety of traditional Chinese medicine prescriptions, providing patients with more personalized and precise treatment plans. Attached Figure Description

[0032] Figure 1 This is a schematic diagram of the structure of the TCM prescription generation system based on data analysis according to an embodiment of the present invention;

[0033] Figure 2 This is a logic diagram of the preliminary labeling and classification of all feature individuals in an embodiment of the present invention;

[0034] Figure 3 A logic diagram for determining the final combination in an embodiment of the present invention. Detailed Implementation

[0035] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0036] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0037] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0038] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0039] This invention provides a data analysis-based traditional Chinese medicine prescription generation system, comprising:

[0040] The information acquisition module is used to acquire a feature description set of the analysis target, wherein the feature description set includes sensitive sub-features, general sub-features and quantifiable sub-features;

[0041] The database module stores the complete feature description set corresponding to each feature individual, the specific data of the quantifiable sub-features in each feature description set, and the status label corresponding to each feature description set; and records the final combination corresponding to the complete feature description set of each feature individual, wherein the final combination includes the final component and the corresponding component content;

[0042] The data integration module, which is connected to the database module, is used to perform basic matching on all the feature description sets to classify all feature individuals into preliminary labels, and to analyze the quantifiable sub-features in each preliminary label category to determine the feature similarity type, and to combine the state labels to integrate the final combination of the corresponding feature similarity types into a state storage type.

[0043] The output module is connected to the information acquisition module and the data integration module respectively, and is used to analyze the feature description set of the analysis target to obtain an initial combination, and output the final combination based on the analysis results of the initial combination;

[0044] The model analysis module, which is connected to the data integration module and the output module respectively, is used to adjust the initial combination based on the pharmacological database and the sensitivity sub-features of the analysis target.

[0045] In practice, both the initial and final combinations are traditional Chinese medicine formulas. The feature description set records the target's symptoms. Sensitive sub-features are symptoms that cannot be used or encountered, such as allergens of the target. Quantifiable sub-features are features that can be quantified by explicit data, such as the number of coughs per hour, heart rate, and body temperature. General sub-features are non-quantifiable sub-features in the symptoms, such as nasal congestion, pain, and body sensation.

[0046] The database module uses a graph database or graph data structure to store data. Feature individuals, feature description sets, state labels, final components, and component contents are all stored as nodes or edges in the graph to facilitate efficient data retrieval and analysis using graph algorithms.

[0047] When performing quantified sub-feature analysis, the data integration module can utilize node similarity comparison methods in graph algorithms, such as graph embedding, to calculate the similarity between individuals with different features within the graph structure, thereby more accurately determining the type of feature similarity. Furthermore, other graph algorithms, such as community detection algorithms, can be used to identify groups of individuals with similar symptoms.

[0048] This invention provides a comprehensive understanding of a patient's specific condition by acquiring a feature description set of the analysis target, including sensitive sub-features, general sub-features, and quantifiable sub-features. The database module stores a rich collection of individual features and their corresponding feature description sets and status labels, providing a solid data foundation for subsequent analysis. The data integration module performs basic matching and analysis on the feature description sets, and preliminary label classification improves the accuracy and efficiency of the analysis. The output module generates initial combinations based on the analysis results and further outputs final combinations, enabling rapid prescription recommendations. The model analysis module adjusts the initial combinations based on a pharmacological database and sensitive sub-features, ensuring the safety and effectiveness of the prescriptions. This invention not only improves the scientific rigor and rationality of TCM prescription generation but also effectively enhances the personalization and precision of TCM prescriptions, providing patients with safer and more effective treatment options, and also strongly supports the modernization and intelligent development of TCM.

[0049] Specifically, the database module stores several feature individuals and their complete feature description sets, and the number of feature description sets is greater than or equal to the total number of feature individuals.

[0050] It is understandable that the status labels stored in the database module are the target's disease name.

[0051] Please see Figure 2 As shown, this is a logic diagram of the preliminary labeling and classification of all feature individuals according to an embodiment of the present invention. The data integration module performs preliminary labeling and classification of all feature individuals based on the general sub-features and the quantifiable sub-features.

[0052] If any of the quantifiable sub-features exceeds the corresponding threshold, and / or the matching result of the general sub-features with the chronic feature database is lower than the corresponding threshold, then the data integration module determines that the category of the preliminary label classification of the feature individual is the acute prescription category.

[0053] If there are quantifiable sub-features that do not exceed the corresponding threshold and the matching result of the general sub-features and the chronic feature database exceeds the corresponding threshold, then the data integration module determines that the category of the preliminary label classification of the feature individual is the chronic prescription category.

[0054] In implementation, the thresholds corresponding to each quantified sub-feature are determined based on the thresholds corresponding to the pathological states. The corresponding thresholds in the chronic feature database are determined by subtracting twice the standard deviation from the average of the matching results corresponding to the chronic prescription category in the historical prescription database; the matching result is the overlap rate.

[0055] Understandably, the threshold of quantifiable sub-features corresponds to the critical value of the pathological state. If it exceeds the threshold, it indicates that the disease may be in an acute phase or that there are significant abnormal pathological indicators, which meets the characteristics of an acute disease. If the overlap rate of the matching results between the general sub-features and the chronic feature database is lower than the corresponding threshold, it indicates that the consistency with the characteristics of a chronic disease is insufficient, and it is more inclined to be acute. Therefore, as long as either the quantifiable sub-feature exceeds the threshold or the consistency with the chronic feature is insufficient, it is determined to be an acute prescription. Conversely, when the quantifiable sub-features do not exceed the threshold, that is, the pathological indicators are relatively stable, and the matching results between the general sub-features and the chronic feature database exceed the threshold and are highly consistent with typical chronic features, it meets the characteristics of a slow course and stable features of a chronic disease, and is therefore determined to be a chronic prescription.

[0056] Specifically, the data integration module analyzes the quantifiable sub-features in each preliminary label category to determine several feature similarity types.

[0057] All quantizable features are divided into several feature similarity types using the corresponding thresholds of each quantizable sub-feature;

[0058] Among them, different feature similarity types have some of the same quantifiable sub-features.

[0059] In implementation, taking the quantifiable sub-features that only include the number of coughs, heart rate, and body temperature as an example, for the low fever feature with a body temperature of 37.1 to 38.0 degrees Celsius, it can include mild cough with a cough frequency of 10 to 20 times and severe cough with a cough frequency of 21 to 30 times (the division method can be any one of the existing technologies, and this is not specifically limited). That is, all the quantifiable sub-features contained in a single feature similarity type are within their respective intervals.

[0060] A single status label can contain more than one type of single feature similarity, corresponding to various degrees of severity of the same type of disease.

[0061] For the final combination in the acute prescription category, extract its high-frequency components and content range, and determine the high-frequency quantifiable sub-features. Then, determine the similar feature type by corresponding the high-frequency quantifiable sub-features and the corresponding components, and store all combinations as the acute state storage type with the corresponding state label.

[0062] For the final combination in the chronic prescription category, extract its core component combination and dosage adjustment method to form a chronic state storage type;

[0063] Specifically, the output module generates an initial combination based on the feature description set of the analysis target:

[0064] If the analysis target is classified as an acute prescription category, the output module matches the similarity between its quantifiable sub-features and each combination in the acute state storage type, selects several final combinations with the highest similarity as candidate combinations, and calculates the component content of each candidate combination by weighting according to the degree of abnormality of the quantifiable sub-features.

[0065] If the analysis target is classified as a chronic prescription category, the output module matches the correlation between its general sub-features and the core component combination in the chronic state storage type, and adjusts the basic component allocation ratio according to the matching results of the chronic feature database.

[0066] In practice, candidate combinations may differ in their formulations. The selection is based on the most frequently occurring types of formulations in the target historical prescriptions. If an initial combination cannot be obtained, all other candidate combinations are excluded, and the combination with the highest similarity is chosen as the initial combination.

[0067] For each symptom feature, two weight values ​​are calculated:

[0068] Quantifiable symptom weight: calculated based on the proportion of the actual measured value of the feature that exceeds the normal threshold. General symptom weight: calculated based on the degree of matching between the feature and the chronic feature database. The comprehensive symptom weight is synthesized from the above two types of weights according to a preset ratio, of which the quantifiable symptom weight accounts for 60% and the general symptom weight accounts for 40%. This weight can be adjusted by the number of the two sub-features, but the final general symptom weight shall not exceed 50%.

[0069] As a preferred embodiment of the present invention:

[0070] Component dosage adjustment:

[0071] For each component in the initial combination, the dosage is adjusted based on the overall weight of its corresponding symptom:

[0072] Acute symptoms: Dosage adjustment range = original dose × overall symptom weight × acute adjustment coefficient

[0073] Chronic conditions: Dosage adjustment range = Original dose × Overall symptom weight × Chronic adjustment coefficient

[0074] The acute adjustment coefficient is set at 0.4, and the chronic adjustment coefficient is set at 0.2.

[0075] Understandably, based on historical data analysis, the standard deviation of data fluctuation for acute symptoms is usually twice that of chronic symptoms. Furthermore, calculations based on the pharmacokinetic cumulative effect model require that the dose fluctuation should not exceed ±20% for long-term medication safety. Moreover, the effective dose adjustment range for acute symptoms is usually within ±40% of the original dose. Exceeding this range may cause adverse reactions. Therefore, the acute adjustment coefficient is set at 0.4, and the chronic adjustment coefficient is set at 0.2.

[0076] Component efficacy screening:

[0077] Calculate the coverage score of each component for the main symptoms:

[0078] Main efficacy contribution value = corresponding symptom weight × main efficacy weight

[0079] Secondary efficacy contribution value = corresponding symptom weight × secondary efficacy weight

[0080] Total coverage score = Contribution value of primary efficacy + Contribution value of secondary efficacy

[0081] Coverage thresholds are set based on disease type: 0.65 for acute diseases and 0.55 for chronic diseases; components with total coverage scores below the corresponding thresholds are eliminated.

[0082] Compatibility optimization: Principal drug enhancement: Increase the weight of the component with the largest dosage by 30%; Adjuvant drug restriction: The total dosage of all adjuvant drugs shall not exceed 40% of the total dosage of the combination; Contraindication handling: For component pairs with contraindications, reduce the weight of both by 80%.

[0083] Specifically, the degree of anomaly of the quantifiable sub-feature is determined by the interval difference between the quantifiable sub-feature of the current analysis target and the corresponding feature similarity type in the state storage type;

[0084] The correlation degree is determined based on the number of general sub-features and the degree of consistency between the combinations in the chronic state storage type.

[0085] In practice, if the current analysis target only uses quantified sub-features A and B, sub-feature A corresponds to the interval [a1, a2], the corresponding square value is [m1, m2], and the analysis target corresponds to the value a', and B corresponds to the interval [b1, b2], the corresponding square value is [n1, n2], and the analysis target corresponds to the value b', then the abnormality degree α of A and B' is (a'-a1) / (a2-a1), and β is (b'-b1) / (b2-b1), then the weight of A is α / α+β.

[0086] Please see Figure 3 As shown, it is a logic diagram for determining the final combination in an embodiment of the present invention. The model analysis module screens each component in the initial combination according to the sensitive sub-features to obtain the screened components, and determines the alternative components according to the pharmacological database.

[0087] The model analysis module obtains a second combination based on the screened components and the alternative components.

[0088] Specifically, the model analysis module determines the rationality of the second combination based on the pharmacological database, and determines the final combination based on the rationality judgment result;

[0089] If the rationality judgment result of the second combination is normal, then the components of the final combination determined by the model analysis module are the same as those of the second combination;

[0090] If the rationality judgment result of the second combination is abnormal, the model analysis module determines to repeat the process of determining alternative components and corresponding rationality based on the pharmacological database.

[0091] The model analysis module screens components in the initial combination based on sensitivity features and identifies alternative components using a pharmacological database, thus obtaining a second combination. This process reflects a precise consideration of individual patient differences, avoiding the negative impact of contraindications such as allergens on treatment efficacy. Simultaneously, by assessing the rationality of the second combination, it ensures that the final combination conforms to pharmacological principles and maximizes therapeutic effects. If the rationality assessment result is normal, the final combination is directly determined to be the same as the second combination, ensuring the stability and reliability of the treatment plan. If an abnormality occurs, the process of identifying alternative components and conducting rationality assessments is repeated to continuously optimize the combination until a reasonable final combination is obtained. This mechanism effectively improves the scientific validity and safety of traditional Chinese medicine prescriptions, providing patients with more personalized and precise treatment plans.

[0092] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using dedicated hardware-based apparatus to perform the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0093] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

[0094] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A data analysis-based traditional Chinese medicine prescription generation system, characterized in that, include: The information acquisition module is used to acquire a feature description set of the analysis target, wherein the feature description set includes sensitive sub-features, general sub-features and quantifiable sub-features; The database module stores the complete feature description set corresponding to each feature individual, the specific data of the quantifiable sub-features in each feature description set, and the status label corresponding to each feature description set; and records the final combination corresponding to the complete feature description set of each feature individual, wherein the final combination includes the final component and the corresponding component content; The data integration module, which is connected to the database module, is used to perform basic matching on all the feature description sets to classify all feature individuals into preliminary labels, and to analyze the quantifiable sub-features in each preliminary label category to determine the feature similarity type, and to combine the state labels to integrate the final combination of the corresponding feature similarity types into a state storage type. The output module is connected to the information acquisition module and the data integration module respectively, and is used to analyze the feature description set of the analysis target to obtain an initial combination, and output the final combination based on the analysis results of the initial combination; The model analysis module, which is connected to the data integration module and the output module respectively, is used to adjust the initial combination according to the pharmacological database and the sensitive sub-features of the analysis target; The data integration module performs preliminary label classification on all feature individuals based on the general sub-features and the quantifiable sub-features. If any of the quantifiable sub-features exceeds the corresponding threshold, and / or the matching result of the general sub-features with the chronic feature database is lower than the corresponding threshold, then the data integration module determines that the category of the preliminary label classification of the feature individual is the acute prescription category. If there are quantifiable sub-features that do not exceed the corresponding threshold and the matching result of the general sub-features and the chronic feature database exceeds the corresponding threshold, then the data integration module determines that the category of the preliminary label classification of the feature individual is the chronic prescription category.

2. The data analysis-based TCM prescription generation system according to claim 1, characterized in that, The database module stores several feature individuals and their complete feature description sets, with the number of feature description sets being greater than or equal to the total number of feature individuals.

3. The data analysis-based TCM prescription generation system according to claim 1, characterized in that, The data integration module analyzes the quantifiable sub-features in each preliminary label category to determine several feature similarity types. All quantizable features are divided into several feature similarity types using the corresponding thresholds of each quantizable sub-feature; Among them, different feature similarity types have some of the same quantifiable sub-features.

4. The data analysis-based TCM prescription generation system according to claim 3, characterized in that, The output module generates an initial combination based on the feature description set of the analysis target: If the analysis target is classified as an acute prescription category, the output module matches the similarity between its quantifiable sub-features and each combination in the acute state storage type, selects several final combinations with the highest similarity as candidate combinations, and calculates the component content of each candidate combination by weighting according to the degree of abnormality of the quantifiable sub-features. If the analysis target is classified as a chronic prescription category, the output module matches the correlation between its general sub-features and the core component combination in the chronic state storage type, and adjusts the basic component allocation ratio according to the matching results of the chronic feature database.

5. The data analysis-based TCM prescription generation system according to claim 4, characterized in that, The degree of anomaly of the quantifiable sub-feature is determined by the interval difference between the quantifiable sub-feature of the current analysis target and the corresponding feature similarity type in the state storage type; The correlation degree is determined based on the number of general sub-features and the degree of consistency between the combinations in the chronic state storage type.

6. The data analysis-based TCM prescription generation system according to claim 5, characterized in that, The model analysis module screens each component in the initial combination based on the sensitive sub-features to obtain the screened components, and determines the alternative components based on the pharmacological database; The model analysis module obtains a second combination based on the screened components and the alternative components.

7. The data analysis-based TCM prescription generation system according to claim 6, characterized in that, The model analysis module determines the rationality of the second combination based on the pharmacological database, and determines the final combination based on the rationality judgment result; If the rationality judgment result of the second combination is normal, then the components of the final combination determined by the model analysis module are the same as those of the second combination; If the rationality judgment result of the second combination is abnormal, the model analysis module determines to repeat the process of determining alternative components and corresponding rationality based on the pharmacological database.

Images