A Traditional Chinese Medicine (TCM) model and system for differentiating syndromes during breast cancer chemotherapy and its construction method.

Abstract

This invention discloses a TCM syndrome differentiation model for breast cancer chemotherapy, including four syndrome types: liver stagnation and yin deficiency, qi and yin deficiency, spleen deficiency and blood deficiency, and qi stagnation and blood stasis, as well as discrimination thresholds; the discrimination threshold for all four syndrome types is 22.5. This invention also relates to a system and method for constructing a TCM syndrome differentiation model for breast cancer chemotherapy. This invention combines TCM with statistics and big data to construct a TCM syndrome differentiation model for breast cancer chemotherapy and establishes a method for constructing the model. The model constructed by this invention standardizes and normalizes the application of TCM in breast cancer chemotherapy, providing a new model for the research, application, and promotion of TCM syndrome differentiation.

Description

Technical Field

[0001] This invention belongs to the field of traditional Chinese medicine, specifically relating to a TCM syndrome differentiation and classification model, system, and construction method for breast cancer chemotherapy. Background Technology

[0002] Breast cancer is a common malignant tumor in women, with 2.08 million new cases worldwide according to the 2018 Global Cancer Database. The application of Traditional Chinese Medicine (TCM) is a distinctive feature and advantage of breast cancer prevention and treatment research in my country. Numerous literature reports have confirmed that TCM intervention can effectively improve the immune status of breast cancer patients, enhance their quality of life, reduce recurrence and metastasis rates, and prolong disease-free survival. Syndrome differentiation and treatment is the core of TCM treatment; objective and reasonable staged syndrome differentiation helps guide clinical prescription and medication, improves treatment efficacy, and leverages the advantages of TCM.

[0003] However, currently, there is a lack of unified diagnostic standards for the TCM (Traditional Chinese Medicine) classification of breast cancer. For example, *Traditional Chinese Surgery* classifies breast cancer into four types: liver qi stagnation, imbalance of the Chong and Ren meridians, heat toxin accumulation, and deficiency of both qi and blood; *Standards for Diagnosis and Efficacy of TCM Syndromes* classifies it into four types: deficiency of vital energy and intense toxicity, liver qi stagnation and phlegm accumulation, and imbalance of the Chong and Ren meridians; *Modern TCM Breast Diseases* classifies breast cancer into four types: phlegm and blood stasis, liver qi stagnation, and imbalance of the Chong and Ren meridians. Thus, different guidebooks and researchers have yielded different classifications of breast cancer, leading to clinical controversy. Therefore, numerous studies have been conducted to analyze and discuss these classifications, aiming to scientifically evaluate the TCM syndrome differentiation of breast cancer. By sorting out the relevant literature on TCM syndrome differentiation of breast cancer published in Chinese journals in recent years, the research situation of TCM syndrome differentiation of breast cancer in recent years can be summarized into the following three types: (1) Traditional TCM syndrome differentiation is a summary of the clinical experience or understanding of an expert. Generally, there is no mandatory requirement for sample size, which may have defects such as non-standard syndrome types and poor clinical guidance; (2) Delphi method: The Delphi method is to solicit expert opinions on the question in an anonymous form, then sort out and summarize the opinions, and then feed them back to the experts. The opinions are repeatedly solicited and the results are revised in order to reach a consensus. Although the Delphi method is still based on expert experience, it is more in line with clinical practice, the results are relatively objective, and the clinical acceptance is higher than that of traditional syndrome differentiation; (3) Statistical syndrome differentiation is a syndrome differentiation study based on the symptoms of real-world patients with a large sample. It uses statistical methods such as cluster analysis, principal component analysis, and factor analysis. Its results are closer to the syndrome differentiation characteristics of real-world patients. Clinical experience is a valuable resource for learning Traditional Chinese Medicine (TCM), but standardized research on TCM syndrome differentiation cannot be accomplished by the experience of one person alone. It requires more objective evidence and continuous revision of previous research to improve the TCM syndrome differentiation research on breast cancer and fully leverage the advantages of TCM. Incorporating statistical methods makes the research more comprehensive, the diagnostic criteria more objective, and compensates for the lack of clinical acceptance in empirical medicine, making the syndrome differentiation standards more scientific and reliable.

[0004] Furthermore, with the advent of the big data era, the establishment and application of mathematical models have provided better analytical capabilities for solving complex problems involving multiple levels and factors in medical clinical research. Recently, some scholars have attempted to construct TCM syndrome models. For example, Huang Zhongyu et al. used structural equation modeling to explore quantitative analysis methods for TCM syndromes in patients with diarrhea-predominant irritable bowel syndrome, thus providing a good solution for the quantitative analysis and application of TCM syndrome theory. Li Wenlin et al. preliminarily analyzed the feasibility of this method in the diagnosis of chronic gastritis syndromes using factor analysis, attempting to deduce the diagnostic rules for gastritis. However, there is currently no statistically based technology for establishing a TCM syndrome differentiation and identification model for breast cancer chemotherapy. Summary of the Invention

[0005] Therefore, it is necessary to provide a TCM syndrome differentiation model for breast cancer chemotherapy to address at least one of the aforementioned technical problems. This model can accurately and quickly determine the TCM syndrome type (hereinafter referred to as syndrome type) of a patient. The model includes four syndrome types:

[0006] The syndrome items and their weight coefficients for Liver Qi Stagnation and Yin Deficiency Syndrome are as follows: Irritability and anger with a weight coefficient of 11, Thin white tongue coating with a weight coefficient of 6, Forgetfulness with a weight coefficient of 6, Chest and hypochondriac distending pain with a weight coefficient of 5, Dry mouth with a weight coefficient of 5, Hot flashes with a weight coefficient of 4, Chest and abdominal pain with a weight coefficient of 4, Bone pain with a weight coefficient of 3, Abdominal distension with a weight coefficient of 3, and Chest tightness with a weight coefficient of 3.

[0007] The syndrome of Qi and Yin deficiency is defined by the following symptoms and their corresponding weight coefficients: deep pulse (18), thin white tongue coating (11), red tongue (11), tidal fever (4), forgetfulness (1), dry mouth (1), pale complexion (1), soreness and weakness of the lower back and knees (1), spontaneous sweating (1), and ruddy complexion (1).

[0008] The corresponding syndrome items and their weight coefficients for the spleen deficiency and blood deficiency syndrome are as follows: pale tongue with a weight coefficient of 13, white and greasy tongue coating with a weight coefficient of 12, pale complexion with a weight coefficient of 7, dry mouth with a weight coefficient of 6, poor appetite with a weight coefficient of 5, thready pulse with a weight coefficient of 4, and swollen tongue with a weight coefficient of 3.

[0009] The syndrome items and their weight coefficients for the syndrome items are as follows: ecchymosis on the tongue and its weight coefficient is 21, cyanosis of the lips and its weight coefficient is 13, varicose veins under the tongue and its weight coefficient is 6, wiry pulse and its weight coefficient is 6, sallow complexion and its weight coefficient is 2, numbness of the limbs and its weight coefficient is 1, and oral ulcers and its weight coefficient is 1.

[0010] It also includes the discrimination thresholds for the four types of evidence; the discrimination threshold for all four types of evidence is 22.5.

[0011] This invention also provides a method for constructing a TCM syndrome differentiation and classification model for breast cancer during chemotherapy, comprising:

[0012] S1. Collection and entry of TCM syndrome information of breast cancer patients undergoing chemotherapy: Information of breast cancer patients undergoing chemotherapy was collected using a TCM syndrome questionnaire for breast cancer chemotherapy; the information included syndrome item information; each syndrome item was statistically analyzed and entered using a binary classification method to establish a TCM syndrome database for breast cancer chemotherapy.

[0013] S2. Establishment of a Traditional Chinese Medicine (TCM) syndrome differentiation model for breast cancer chemotherapy:

[0014] S21. Retrieve the binary classification information from the TCM syndrome database for breast cancer chemotherapy, and perform logistic regression analysis using SPSS software to obtain the original weight coefficients of each syndrome item.

[0015] S22. Plot the ROC curve of each syndrome type according to the original weight coefficients of the syndrome items corresponding to each syndrome type, and obtain the Youden index of each syndrome type at different hypothesis discrimination thresholds; take the hypothesis discrimination threshold corresponding to the maximum Youden index of each syndrome type as the candidate discrimination threshold; take the value with the highest repetition from the candidate discrimination thresholds as the discrimination threshold of each syndrome type.

[0016] S23. Set the total score of each syndrome type to M, calculate the adjustment coefficient F of the weight coefficient and the original weight coefficient of the syndrome item corresponding to each syndrome type using Formula 1; and obtain the weight coefficient of the syndrome item corresponding to each syndrome type using Formula 2; Maximum hypothesis discrimination threshold + 3 ≥ M ≥ the discrimination threshold;

[0017] Formula 1: Sum of the original weight coefficients of the syndrome items corresponding to the syndrome type × F = M

[0018] Formula 2: The weight coefficient of the syndrome item corresponding to the syndrome type is equal to the original weight coefficient of the syndrome item corresponding to the syndrome type × F.

[0019] In this invention, binary classification refers to presence or absence, that is, affirmation or negation, with one of the two being chosen.

[0020] In one or more specific embodiments, binary classification is represented by 0 and 1, where 0 = none and 1 = yes; or conversely, 0 = yes and 1 = none.

[0021] Furthermore, M = 50.

[0022] Furthermore, it also includes:

[0023] S3. Validation of the TCM syndrome differentiation model for breast cancer chemotherapy: Compare the consistency between the syndrome types determined by the TCM syndrome differentiation model for breast cancer chemotherapy and those determined by TCM breast cancer experts; the TCM syndrome differentiation model for breast cancer chemotherapy uses a binary classification of each syndrome item in the TCM syndrome questionnaire for breast cancer chemotherapy patients, and combines the weight coefficients of each syndrome item in the model to calculate the patient's score for each syndrome type; when the score of a certain syndrome type is ≥ the discrimination threshold, the model determines that the patient belongs to that syndrome type; when the scores of multiple syndrome types are all ≥ the discrimination threshold, the syndrome type corresponding to the highest score is taken as the patient's syndrome type.

[0024] In one or more specific embodiments, the TCM syndrome survey form for breast cancer chemotherapy includes an investigation of general symptoms, tongue appearance, and pulse appearance; the general symptoms include fatigue, pale nails, frequent urination, chest tightness, edema of the limbs, difficulty urinating, irritability, dry mouth, loose stools, forgetfulness, insomnia, constipation, anxiety, poor appetite, cyanosis of the lips, hot flashes, bone pain, ruddy complexion, spontaneous sweating, numbness of the limbs, sallow complexion, tinnitus, dizziness, dull complexion, dry eyes, abdominal distension, pale complexion, chest and rib pain, oral ulcers, soreness of the waist and knees, nausea and vomiting, chest and abdominal pain, and aversion to cold.

[0025] Furthermore, tongue appearance includes tongue color, tongue shape, and tongue coating.

[0026] Furthermore, tongue color includes pale red, pale white, red, purple, and dark.

[0027] Furthermore, tongue shape includes teeth marks, enlargement, cracks, thinness, prickles, varicose veins under the tongue, and sublingual ecchymosis.

[0028] Furthermore, tongue coating can be categorized as thin white, thin yellow, white and greasy, yellow and greasy, or smooth and peeling.

[0029] Furthermore, pulse characteristics include deep, thin, wiry, slippery, hesitant, rapid, and weak.

[0030] In one or more specific embodiments, the TCM syndrome survey form for breast cancer chemotherapy includes an investigation of the following syndrome items: irritability, thin white tongue coating, forgetfulness, chest and rib pain, dry mouth, hot flashes, chest and abdominal pain, bone pain, abdominal distension, chest tightness, deep pulse, red tongue, pale complexion, soreness and weakness of the waist and knees, spontaneous sweating, rosy complexion, pale white tongue, white and greasy tongue coating, poor appetite, thready pulse, enlarged tongue, ecchymosis on the tongue, cyanosis of the lips, varicose veins under the tongue, wiry pulse, sallow complexion, numbness of the limbs, and oral ulcers.

[0031] This invention also provides a method for constructing the TCM syndrome differentiation and classification model for breast cancer during chemotherapy as described above, comprising:

[0032] S1. Collection and entry of TCM syndrome information of breast cancer patients undergoing chemotherapy: Information of breast cancer patients undergoing chemotherapy was collected using a TCM syndrome questionnaire for breast cancer chemotherapy; the information included syndrome item information; each syndrome item was statistically analyzed and entered using a binary classification method to establish a TCM syndrome database for breast cancer chemotherapy.

[0033] S2. Establishment of a Traditional Chinese Medicine (TCM) syndrome differentiation model for breast cancer chemotherapy:

[0034] S21. Retrieve the binary classification information from the TCM syndrome database for breast cancer chemotherapy, and perform logistic regression analysis using SPSS software to obtain the original weight coefficients of each syndrome item.

[0035] S22. Plot the ROC curve of each syndrome type according to the original weight coefficients of the syndrome items corresponding to each syndrome type, and obtain the Youden index of each syndrome type at different hypothesis discrimination thresholds; take the hypothesis discrimination threshold corresponding to the maximum Youden index of each syndrome type as the candidate discrimination threshold; take the value with the highest repetition from the candidate discrimination thresholds as the discrimination threshold of each syndrome type.

[0036] S23. Set the total score of each syndrome type to M, calculate the adjustment coefficient F of the weight coefficient and the original weight coefficient of the syndrome item corresponding to each syndrome type using Formula 1; and obtain the weight coefficient of the syndrome item corresponding to each syndrome type using Formula 2; Maximum hypothesis discrimination threshold + 3 ≥ M ≥ the discrimination threshold;

[0037] Formula 1: Sum of the original weight coefficients of the syndrome items corresponding to the syndrome type × F = M

[0038] Formula 2: The weight coefficient of the syndrome item corresponding to the syndrome type is equal to the original weight coefficient of the syndrome item corresponding to the syndrome type × F.

[0039] In one or more specific embodiments, binary classification is represented by 0 and 1, where 0 = none and 1 = yes; or conversely, 0 = yes and 1 = none.

[0040] Furthermore, M = 50.

[0041] Furthermore, it also includes:

[0042] S3. Validation of the TCM syndrome differentiation model for breast cancer chemotherapy: Compare the consistency between the syndrome types determined by the TCM syndrome differentiation model for breast cancer chemotherapy and those determined by TCM breast cancer experts; the TCM syndrome differentiation model for breast cancer chemotherapy uses a binary classification of each syndrome item in the TCM syndrome questionnaire for breast cancer chemotherapy patients, and combines the weight coefficients of each syndrome item in the model to calculate the patient's score for each syndrome type; when the score of a certain syndrome type is ≥ the discrimination threshold, the model determines that the patient belongs to that syndrome type; when the scores of multiple syndrome types are all ≥ the discrimination threshold, the syndrome type corresponding to the highest score is taken as the patient's syndrome type.

[0043] In one or more specific embodiments, the TCM syndrome survey form for breast cancer chemotherapy includes an investigation of general symptoms, tongue appearance, and pulse appearance; the general symptoms include fatigue, pale nails, frequent urination, chest tightness, edema of the limbs, difficulty urinating, irritability, dry mouth, loose stools, forgetfulness, insomnia, constipation, anxiety, poor appetite, cyanosis of the lips, hot flashes, bone pain, ruddy complexion, spontaneous sweating, numbness of the limbs, sallow complexion, tinnitus, dizziness, dull complexion, dry eyes, abdominal distension, pale complexion, chest and rib pain, oral ulcers, soreness of the waist and knees, nausea and vomiting, chest and abdominal pain, and aversion to cold.

[0044] Furthermore, tongue appearance includes tongue color, tongue shape, and tongue coating.

[0045] Furthermore, tongue color includes pale red, pale white, red, purple, and dark.

[0046] Furthermore, tongue shape includes teeth marks, enlargement, cracks, thinness, prickles, varicose veins under the tongue, and sublingual ecchymosis.

[0047] Furthermore, tongue coating can be categorized as thin white, thin yellow, white and greasy, yellow and greasy, or smooth and peeling.

[0048] Furthermore, pulse characteristics include deep, thin, wiry, slippery, hesitant, rapid, and weak.

[0049] In one or more specific embodiments, the TCM syndrome survey form for breast cancer chemotherapy includes an investigation of the following syndrome items: irritability, thin white tongue coating, forgetfulness, chest and rib pain, dry mouth, hot flashes, chest and abdominal pain, bone pain, abdominal distension, chest tightness, deep pulse, red tongue, pale complexion, soreness and weakness of the waist and knees, spontaneous sweating, rosy complexion, pale white tongue, white and greasy tongue coating, poor appetite, thready pulse, enlarged tongue, ecchymosis on the tongue, cyanosis of the lips, varicose veins under the tongue, wiry pulse, sallow complexion, numbness of the limbs, and oral ulcers.

[0050] This invention also provides a system for constructing a TCM syndrome differentiation and discrimination model for breast cancer during chemotherapy, comprising:

[0051] The information collection and entry module is used to collect information from breast cancer patients undergoing chemotherapy using a Traditional Chinese Medicine (TCM) syndrome survey form for breast cancer chemotherapy. The information includes syndrome item information. Each syndrome item is statistically analyzed and entered using a binary classification method to establish a TCM syndrome database for breast cancer chemotherapy.

[0052] The original weight coefficient acquisition module is used to retrieve binary information from the TCM syndrome database during breast cancer chemotherapy and obtain the original weight coefficients of each syndrome item by performing logistic regression analysis using SPSS software.

[0053] The discrimination threshold acquisition module is used to draw ROC curves for each syndrome type based on the original weight coefficients of the syndrome items corresponding to each syndrome type, and to obtain the Youden index of each syndrome type at different hypothetical discrimination thresholds; to take the hypothetical discrimination threshold corresponding to the maximum Youden index of each syndrome type as the candidate discrimination threshold; and to take the value with the highest repetition from the candidate discrimination thresholds as the discrimination threshold of each syndrome type.

[0054] The weight coefficient generation module is used to set the total score of each syndrome type to M, calculate the weight coefficient of the syndrome item corresponding to each syndrome type and the adjustment coefficient F of the original weight coefficient using formula 1; and obtain the weight coefficient of the syndrome item corresponding to each syndrome type using formula 2; the maximum hypothesis discrimination threshold + 3 ≥ M ≥ the discrimination threshold;

[0055] Formula 1: Sum of the original weight coefficients of the syndrome items corresponding to the syndrome type × F = M

[0056] Formula 2: The weight coefficient of the syndrome item corresponding to the syndrome type is equal to the original weight coefficient of the syndrome item corresponding to the syndrome type × F.

[0057] In one or more specific embodiments, binary classification is represented by 0 and 1, where 0 = none and 1 = yes; or conversely, 0 = yes and 1 = none.

[0058] Furthermore, M = 50.

[0059] Furthermore, it also includes:

[0060] The verification module is used to compare the consistency between the syndrome differentiation and classification model for breast cancer chemotherapy and the syndrome differentiation determined by traditional Chinese medicine breast cancer experts. The traditional Chinese medicine syndrome differentiation and classification model for breast cancer chemotherapy uses a questionnaire on the syndrome differentiation of breast cancer chemotherapy patients to classify each syndrome item into two categories, and combines the weight coefficients of each syndrome item in the model to calculate the patient's score for each syndrome type. When the score of a certain syndrome type is greater than or equal to the discrimination threshold, the model determines that the patient belongs to that syndrome type. When the scores of multiple syndrome types are greater than or equal to the discrimination threshold, the syndrome type corresponding to the highest score is taken as the patient's syndrome type.

[0061] In one or more specific embodiments, the TCM syndrome survey form for breast cancer chemotherapy includes an investigation of general symptoms, tongue appearance, and pulse appearance; the general symptoms include fatigue, pale nails, frequent urination, chest tightness, edema of the limbs, difficulty urinating, irritability, dry mouth, loose stools, forgetfulness, insomnia, constipation, anxiety, poor appetite, cyanosis of the lips, hot flashes, bone pain, ruddy complexion, spontaneous sweating, numbness of the limbs, sallow complexion, tinnitus, dizziness, dull complexion, dry eyes, abdominal distension, pale complexion, chest and rib pain, oral ulcers, soreness of the waist and knees, nausea and vomiting, chest and abdominal pain, and aversion to cold.

[0062] Furthermore, tongue appearance includes tongue color, tongue shape, and tongue coating.

[0063] Furthermore, tongue color includes pale red, pale white, red, purple, and dark.

[0064] Furthermore, tongue shape includes teeth marks, enlargement, cracks, thinness, prickles, varicose veins under the tongue, and sublingual ecchymosis.

[0065] Furthermore, tongue coating can be categorized as thin white, thin yellow, white and greasy, yellow and greasy, or smooth and peeling.

[0066] Furthermore, pulse characteristics include deep, thin, wiry, slippery, hesitant, rapid, and weak.

[0067] In one or more specific embodiments, the TCM syndrome survey form for breast cancer chemotherapy includes an investigation of the following syndrome items: irritability, thin white tongue coating, forgetfulness, chest and rib pain, dry mouth, hot flashes, chest and abdominal pain, bone pain, abdominal distension, chest tightness, deep pulse, red tongue, pale complexion, soreness and weakness of the waist and knees, spontaneous sweating, rosy complexion, pale white tongue, white and greasy tongue coating, poor appetite, thready pulse, enlarged tongue, ecchymosis on the tongue, cyanosis of the lips, varicose veins under the tongue, wiry pulse, sallow complexion, numbness of the limbs, and oral ulcers.

[0068] This invention also provides a TCM syndrome differentiation and identification system for breast cancer chemotherapy, which includes the TCM syndrome differentiation and identification model for breast cancer chemotherapy as described above.

[0069] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the construction method described above.

[0070] This invention combines Traditional Chinese Medicine (TCM) with statistics and big data to construct a TCM syndrome differentiation model for breast cancer chemotherapy and establish a method for model construction. This invention analyzes the correlation between TCM syndrome differentiation and clinical factors, providing a theoretical basis for the application of TCM in breast cancer chemotherapy. Furthermore, the model constructed in this invention standardizes and normalizes the application of TCM in breast cancer chemotherapy, providing a new model for the research, application, and promotion of TCM syndrome differentiation. The model, construction method, system, and media involved in this invention avoid subjective errors by doctors in TCM syndrome differentiation of breast cancer chemotherapy patients, and are particularly helpful for TCM syndrome differentiation in areas with poor medical conditions and limited medical resources.

[0071] The following will further explain the concept, specific structure, and technical effects of the present invention in conjunction with the accompanying drawings, so as to fully understand the purpose, features, and effects of the present invention. Attached Figure Description

[0072] Figure 1 This is a flowchart illustrating the construction method of the TCM syndrome differentiation and classification model for breast cancer chemotherapy in a specific implementation.

[0073] Figure 2 This is a questionnaire for investigating the symptoms of breast cancer using traditional Chinese medicine, as described in a specific implementation plan.

[0074] Figure 3 This is the ROC curve of the four syndrome types in the TCM syndrome differentiation of breast cancer during chemotherapy in Example 2. Detailed Implementation

[0075] To make the technical means, inventive features, objectives, and effects of the invention readily understandable, the invention is further illustrated below with reference to specific figures. However, the invention is not limited to the embodiments described below.

[0076] It should be noted that the structures, proportions, sizes, etc., illustrated in the accompanying drawings of this specification are only used to complement the content disclosed in the specification for those skilled in the art to understand and read, and are not intended to limit the conditions under which the present invention can be implemented. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size, without affecting the effects and objectives that the present invention can produce, should still fall within the scope of the technical content disclosed in the present invention.

[0077] The construction process of the TCM syndrome differentiation and classification model for breast cancer chemotherapy is as follows, see below. Figure 1 :

[0078] Five hundred eligible patients undergoing adjuvant chemotherapy for breast cancer were selected for enrollment based on exclusion criteria. Of these, 400 were included in the training set and 100 in the test set. In the training set, a questionnaire survey was conducted using the Traditional Chinese Medicine (TCM) syndrome differentiation questionnaire for breast cancer. The TCM syndrome information from the questionnaires was then extracted to establish a TCM syndrome differentiation model for breast cancer undergoing chemotherapy.

[0079] In the test set, a questionnaire survey was conducted using the same TCM syndrome questionnaire for breast cancer. Then, the TCM syndrome information from the questionnaires was extracted, and the model built from the training set was used to determine the syndrome types of the patients in the test set. The results of the model determination were compared with the results of manual determination to assess the degree of agreement. In this specific embodiment, manual determination involved two TCM breast cancer experts determining the syndrome types of the test set based on the TCM syndrome questionnaire for breast cancer completed by 100 patients.

[0080] Example 1: Collection and Input of Traditional Chinese Medicine Syndrome Information for Breast Cancer Patients Undergoing Chemotherapy

[0081] 1) Case source: A total of 500 patients with malignant breast tumors who underwent chemotherapy at Yueyang Hospital of Integrated Traditional Chinese and Western Medicine affiliated to Shanghai University of Traditional Chinese Medicine and Shanghai Fudan University Cancer Hospital were collected from December 2019 to January 2021. 400 cases were used as the experimental set for model construction and 100 cases were used as the validation set for model validation.

[0082] 2) Diagnostic criteria: Refer to "Surgery" (Chief Editor: Chen Xiaoping, Beijing: People's Medical Publishing House, 2nd edition, August 2010). Based on the patient's clinical manifestations, physical examination, imaging examinations, and pathological examinations, the diagnosis was malignant breast tumor.

[0083] 3) Inclusion criteria: ① Female; ② Breast cancer patients with a clear pathological diagnosis; ③ Breast cancer-related surgical resection has been performed; ④ Currently receiving adjuvant chemotherapy for breast cancer; ⑤ Complete relevant medical history and examination data; ⑥ Informed consent has been signed.

[0084] 4) Exclusion criteria: ① Those who cannot cooperate in completing the questionnaire, including those with language difficulties or mental illnesses; ② Those with advanced breast cancer who have undergone multiple lines of chemotherapy; ③ Patients undergoing neoadjuvant chemotherapy.

[0085] 5) Collection and entry of TCM syndrome information during breast cancer chemotherapy: Data was collected from the second week after patients received adjuvant chemotherapy. General clinical information and TCM syndrome information were recorded according to the "Breast Cancer Chemotherapy Period TCM Syndrome Survey Form." Each syndrome element was statistically analyzed using a binary classification (0 = none, 1 = present). A corresponding data entry program was established using EpiData database software. The database was tested and data entry was completed, establishing a dedicated "Breast Cancer Chemotherapy Period TCM Syndrome Database" for this study. This study employed a fixed, dedicated data collection team, avoiding leading language. Tongue and pulse examinations were collected jointly by two physicians. Data was entered twice by two individuals, and the results were verified to minimize entry errors.

[0086] Example 2: Establishment of a Traditional Chinese Medicine (TCM) syndrome differentiation model for breast cancer chemotherapy.

[0087] 1) SPSS 21.0 software was used to perform statistical analysis on the data in the TCM syndrome database for breast cancer chemotherapy. Four syndrome types were used for TCM differentiation: Liver Qi Deficiency with Yin Deficiency, Qi and Yin Deficiency, Spleen Deficiency with Blood Deficiency, and Qi Stagnation with Blood Stasis. The original weight coefficients corresponding to each syndrome item in the TCM syndrome differentiation relationship for breast cancer chemotherapy (see Tables 1-4) were used as the scoring ratios. Tables 1-4 show the results of the four-class logistic regression analysis (retaining syndrome items with original weight coefficients > 0), where B represents the regression coefficient and p represents the probability.

[0088] Table 1. Results of logistic regression analysis for liver stagnation and yin deficiency type

[0089]

[0090]

[0091] Table 2. Results of logistic regression analysis for Qi and Yin deficiency type

[0092] Syndrome Entries B Standard error Wald Degrees of freedom p Original weight coefficients Thin white tongue coating 1.673 0.375 19.86 1 0 5.326 forgetful -1.118 0.344 10.576 1 0.001 0.327 Dry mouth -1.295 0.336 14.852 1 0 0.274 pale -0.948 0.382 6.166 1 0.013 0.387 Hot flashes 0.754 0.358 4.437 1 0.035 2.125 Lower back and knee pain -1.419 0.371 14.617 1 0 0.242 spontaneous sweating -1.083 0.364 8.852 1 0.003 0.339 Deep pulse 2.18 0.458 22.666 1 0 8.85 rosy complexion -1.238 0.49 6.392 1 0.011 0.29 Red tongue 1.695 0.483 12.296 1 0 5.444

[0093] Table 3. Results of logistic regression analysis for spleen deficiency and blood deficiency type

[0094] Syndrome Entries B Standard error Wald Degrees of freedom p Original weight coefficients Dry mouth 1.318 0.344 14.661 1 0 3.735 pale 1.488 0.331 20.199 1 0 4.43 Thready pulse 1.046 0.333 9.848 1 0.002 2.847 Nacha 1.185 0.346 11.74 1 0.001 3.271 Pale tongue 2.162 0.345 39.194 1 0 8.687 chubby tongue 0.79 0.349 5.141 1 0.023 2.204 White and greasy tongue coating 2.092 0.437 22.907 1 0 8.104

[0095] Table 4. Results of logistic regression analysis for Qi stagnation and blood stasis type

[0096]

[0097]

[0098] 2) Draw the ROC curve for each syndrome type based on the original weight coefficients of the corresponding syndrome items. (See figure) Figure 3The ROC curve results show that the area under the curve (AUC) for liver qi stagnation and yin deficiency syndrome is 0.95; for qi and yin deficiency syndrome it is 0.715; for spleen deficiency and blood deficiency syndrome it is 0.85; and for qi stagnation and blood stasis syndrome it is 0.975. Since the AUC is greater than 0.7, the model fit meets clinical requirements, indicating that the model has good predictive power for the target data. P < 0.001 indicates that the model is statistically significant (see Table 5).

[0099] Table 5. Area under the ROC curve

[0100]

[0101] 3) The values ​​corresponding to each point on the ROC curve include sensitivity, 1-specificity, hypothesis discrimination threshold, and Youden's index, as shown in Table 6-9. Sensitivity, as the standard for evaluating the probability of detecting a positive value (true positive), is inversely proportional to the false negative rate; specificity, as the standard for evaluating the probability of detecting a negative value (true negative), is inversely proportional to the false negative rate. The Youden's index can be calculated from sensitivity and specificity: Youden's index = sensitivity - (1-specificity). As one method for evaluating the validity of a test, the highest Youden's index (i.e., the maximum Youden's index) indicates high sensitivity and specificity. The hypothesis discrimination threshold corresponding to the maximum Youden's index is used as the candidate discrimination threshold for syndrome differentiation. The value with the highest repeatability among the candidate discrimination thresholds is then selected as the discrimination threshold for the four syndrome types. In this embodiment, the hypothetical discrimination thresholds corresponding to the maximum Youden index for the four types of evidence are 23.5, 21.5, 22.5, and 22.5, respectively. Among them, 22.5 appears twice, i.e., the repetition degree is 2; 23.5 and 21.5 each appear once, i.e., the repetition degree is 1. Therefore, 22.5, which has the highest repetition degree, is taken as the discrimination threshold for the four types of evidence.

[0102] Table 6. Data table corresponding to ROC curves of liver stagnation and yin deficiency syndrome.

[0103]

[0104]

[0105]

[0106] Table 7. Data Table of ROC Curves for Qi and Yin Deficiency Syndrome

[0107]

[0108]

[0109] Table 8. Data table corresponding to ROC curves for spleen deficiency and blood deficiency syndrome.

[0110]

[0111]

[0112]

[0113] Table 9. Data Table of ROC Curves for Qi Stagnation and Blood Stasis Syndrome

[0114]

[0115]

[0116] To facilitate calculation and comparison, the sum of the weight coefficients of all syndrome items corresponding to each syndrome type is set to 50. The weight coefficient of the corresponding syndrome item is calculated using the original weight coefficients in Table 1-4. Taking Liver Qi Stagnation and Yin Deficiency Syndrome as an example, its calculation method is: (4.534+4.954+3.952+3.149+2.481+2.398+2.71+8.596+3.423+4.335)×F1=50. The calculated F1=1.23, where F is the adjustment coefficient between the weight coefficient of the syndrome item corresponding to the syndrome type and the original weight coefficient, and F1 is the adjustment coefficient between the weight coefficient of the syndrome item corresponding to Liver Qi Stagnation and Yin Deficiency Syndrome and the original weight coefficient. The weight coefficient of a syndrome item = the original weight coefficient of the syndrome item × F. Substituting the value of F and rounding it to the nearest integer, we obtain the weight coefficients of the syndrome items (see the values ​​in parentheses of the syndrome items in Table 10), thus obtaining the TCM syndrome differentiation and discrimination model for breast cancer chemotherapy, as shown in Table 10.

[0117] Table 10. Traditional Chinese Medicine (TCM) Syndrome Differentiation Model for Breast Cancer Chemotherapy Period

[0118]

[0119]

[0120] Example 3: Validation of the Traditional Chinese Medicine (TCM) syndrome differentiation model for breast cancer chemotherapy.

[0121] The 100 patients in the validation set included in Example 1 were compared with the TCM syndrome differentiation model for breast cancer chemotherapy established in Example 2, and the TCM syndrome differentiation determined by experts. Using expert diagnosis as the gold standard, the sensitivity, specificity, and accuracy of the TCM syndrome differentiation model for breast cancer chemotherapy established in Example 2 relative to expert diagnosis were calculated. The results are as follows: ① Liver Qi stagnation and Yin deficiency syndrome: Sensitivity: 0.70, Specificity: 0.91, Accuracy: 0.87, Youden index: 0.66; ② Qi and Yin deficiency syndrome: Sensitivity: 0.65, Specificity: 0.95, Accuracy: 0.84, Youden index: 0.86; ③ Spleen deficiency and blood deficiency syndrome: Sensitivity: 0.96, Specificity: 0.92, Accuracy: 0.91, Youden index: 0.63; ④ Qi stagnation and blood stasis syndrome: Sensitivity: 0.95, Specificity: 0.96, Accuracy: 0.96, Youden index: 0.90 (see Table 11).

[0122] Table 11. Validation of the Traditional Chinese Medicine (TCM) syndrome differentiation model for breast cancer chemotherapy.

[0123]

[0124] Note: a = True positive rate (both the model and experts identify this syndrome type); b = False positive rate (the model identifies this syndrome type, but experts do not); c = False negative rate (the model determines it is not this syndrome type, but experts determine it is); d = True negative rate (both the model and experts determine it is not this syndrome type). The sensitivity, specificity, and accuracy of the discrimination model in clinical validation are calculated using a four-fold table, where sensitivity = a / a+c; specificity = d / b+d; and accuracy = a+d / a+b+c+d.

[0125] The preferred embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection defined by the claims.

Claims

1. A breast cancer chemotherapy period TCM syndrome differentiation typing discrimination model, characterized in that, It includes four types of certificates, namely: The syndrome items and their weight coefficients for Liver Qi Stagnation and Yin Deficiency Syndrome are as follows: Irritability and anger with a weight coefficient of 11, Thin white tongue coating with a weight coefficient of 6, Forgetfulness with a weight coefficient of 6, Chest and hypochondriac distending pain with a weight coefficient of 5, Dry mouth with a weight coefficient of 5, Hot flashes with a weight coefficient of 4, Chest and abdominal pain with a weight coefficient of 4, Bone pain with a weight coefficient of 3, Abdominal distension with a weight coefficient of 3, and Chest tightness with a weight coefficient of 3. The syndrome of Qi and Yin deficiency is defined by the following symptoms and their corresponding weight coefficients: deep pulse (18), thin white tongue coating (11), red tongue (11), tidal fever (4), forgetfulness (1), dry mouth (1), pale complexion (1), soreness and weakness of the lower back and knees (1), spontaneous sweating (1), and ruddy complexion (1). The corresponding syndrome items and their weight coefficients for the spleen deficiency and blood deficiency syndrome are as follows: pale tongue with a weight coefficient of 13, white and greasy tongue coating with a weight coefficient of 12, pale complexion with a weight coefficient of 7, dry mouth with a weight coefficient of 6, poor appetite with a weight coefficient of 5, thready pulse with a weight coefficient of 4, and swollen tongue with a weight coefficient of 3. The syndrome items and their weight coefficients for the syndrome items are as follows: ecchymosis on the tongue and its weight coefficient is 21, cyanosis of the lips and its weight coefficient is 13, varicose veins under the tongue and its weight coefficient is 6, wiry pulse and its weight coefficient is 6, sallow complexion and its weight coefficient is 2, numbness of the limbs and its weight coefficient is 1, and oral ulcers and its weight coefficient is 1. It also includes the discrimination thresholds for the four types of evidence; the discrimination threshold for all four types of evidence is 22.

5.

2. A method for constructing a TCM syndrome differentiation type discrimination model for breast cancer chemotherapy, characterized in that, include: S1. Collection and entry of TCM syndrome information of breast cancer patients undergoing chemotherapy: Information of breast cancer patients undergoing chemotherapy was collected using a TCM syndrome questionnaire. The information included syndrome items. Each syndrome item was statistically analyzed and entered using a binary classification method to establish a TCM syndrome database for breast cancer patients undergoing chemotherapy. The TCM syndrome questionnaire for breast cancer patients undergoing chemotherapy included the following syndrome items: irritability, thin white tongue coating, forgetfulness, chest and rib pain, dry mouth, hot flashes, chest and abdominal pain, bone pain, abdominal distension, chest tightness, deep pulse, red tongue, pale complexion, soreness and weakness of the waist and knees, spontaneous sweating, rosy complexion, pale tongue, white and greasy tongue coating, poor appetite, thready pulse, enlarged tongue, ecchymosis on the tongue, cyanosis of the lips, varicose veins under the tongue, wiry pulse, sallow complexion, numbness of the limbs, and oral ulcers. S2. Establishment of a Traditional Chinese Medicine (TCM) syndrome differentiation model for breast cancer chemotherapy: S21. Retrieve the binary classification information from the TCM syndrome database for breast cancer chemotherapy, and perform logistic regression analysis using SPSS software to obtain the original weight coefficients of each syndrome item. S22. Plot the ROC curve of each syndrome type according to the original weight coefficients of the syndrome items corresponding to each syndrome type, and obtain the Youden index of each syndrome type at different hypothesis discrimination thresholds; take the hypothesis discrimination threshold corresponding to the maximum Youden index of each syndrome type as the candidate discrimination threshold; take the value with the highest repetition from the candidate discrimination thresholds as the discrimination threshold of each syndrome type. S23. Set the total score of each syndrome type to M, calculate the adjustment coefficient F of the weight coefficient and the original weight coefficient of the syndrome item corresponding to each syndrome type using Formula 1; and obtain the weight coefficient of the syndrome item corresponding to each syndrome type using Formula 2; Maximum hypothesis discrimination threshold + 3 ≥ M ≥ the discrimination threshold; Formula 1: Sum of the original weight coefficients of the syndrome items corresponding to the syndrome type × F = M Formula 2: The weight coefficient of the syndrome item corresponding to the syndrome type is equal to the original weight coefficient of the syndrome item corresponding to the syndrome type × F.

3. The breast cancer chemotherapy period TCM syndrome differentiation typing discrimination model construction method of claim 2, characterized in that, M=50。 4. The method for constructing the TCM syndrome differentiation and classification model for breast cancer chemotherapy as described in claim 2, characterized in that, Also includes: S3. Validation of the TCM syndrome differentiation model for breast cancer chemotherapy: Compare the consistency between the syndrome types determined by the TCM syndrome differentiation model for breast cancer chemotherapy and those determined by TCM breast cancer experts; the TCM syndrome differentiation model for breast cancer chemotherapy uses a binary classification of each syndrome item in the TCM syndrome questionnaire for breast cancer chemotherapy patients, and combines the weight coefficients of each syndrome item in the model to calculate the patient's score for each syndrome type; when the score of a certain syndrome type is ≥ the discrimination threshold, the model determines that the patient belongs to that syndrome type; when the scores of multiple syndrome types are all ≥ the discrimination threshold, the syndrome type corresponding to the highest score is taken as the patient's syndrome type.

5. A method for constructing a TCM syndrome differentiation and discrimination model for breast cancer chemotherapy as described in claim 1, characterized in that, include: S1. Collection and entry of TCM syndrome information of breast cancer patients undergoing chemotherapy: Information of breast cancer patients undergoing chemotherapy was collected using a TCM syndrome questionnaire for breast cancer chemotherapy; the information included syndrome item information; each syndrome item was statistically analyzed and entered using a binary classification method to establish a TCM syndrome database for breast cancer chemotherapy. S2. Establishment of a Traditional Chinese Medicine (TCM) syndrome differentiation model for breast cancer chemotherapy: S21. Retrieve the binary classification information from the TCM syndrome database for breast cancer chemotherapy, and perform logistic regression analysis using SPSS software to obtain the original weight coefficients of each syndrome item. S22. Plot the ROC curve of each syndrome type according to the original weight coefficients of the syndrome items corresponding to each syndrome type, and obtain the Youden index of each syndrome type at different hypothesis discrimination thresholds; take the hypothesis discrimination threshold corresponding to the maximum Youden index of each syndrome type as the candidate discrimination threshold; take the value with the highest repetition from the candidate discrimination thresholds as the discrimination threshold of each syndrome type. S23. Set the total score of each syndrome type to M, calculate the adjustment coefficient F of the weight coefficient and the original weight coefficient of the syndrome item corresponding to each syndrome type using Formula 1; and obtain the weight coefficient of the syndrome item corresponding to each syndrome type using Formula 2; maximum hypothesis discrimination threshold + 3 ≥ M ≥ the discrimination threshold; M = 50; Formula 1: Sum of the original weight coefficients of the syndrome items corresponding to the syndrome type × F = M Formula 2: The weight coefficient of the syndrome item corresponding to the syndrome type is equal to the original weight coefficient of the syndrome item corresponding to the syndrome type × F.

6. A system for constructing a TCM syndrome differentiation and discrimination model for breast cancer during chemotherapy, characterized in that, include: The information collection and input module is used to collect information from breast cancer patients undergoing chemotherapy using a Traditional Chinese Medicine (TCM) syndrome questionnaire. The information includes syndrome entries. Each syndrome entry is statistically analyzed and entered using a binary classification system to establish a TCM syndrome database for breast cancer patients undergoing chemotherapy. The TCM syndrome questionnaire for breast cancer patients undergoing chemotherapy includes the following syndrome entries: irritability, thin white tongue coating, forgetfulness, chest and rib pain, dry mouth, hot flashes, chest and abdominal pain, bone pain, abdominal distension, chest tightness, deep pulse, red tongue, pale complexion, soreness and weakness of the lower back and knees, spontaneous sweating, rosy complexion, pale tongue, white and greasy tongue coating, poor appetite, thready pulse, enlarged tongue, ecchymosis on the tongue, cyanosis of the lips, varicose veins under the tongue, wiry pulse, dull complexion, numbness of the limbs, and oral ulcers. The original weight coefficient acquisition module is used to retrieve binary information from the TCM syndrome database during breast cancer chemotherapy and obtain the original weight coefficients of each syndrome item by performing logistic regression analysis using SPSS software. The discrimination threshold acquisition module is used to draw ROC curves for each syndrome type based on the original weight coefficients of the syndrome items corresponding to each syndrome type, and to obtain the Youden index of each syndrome type at different hypothetical discrimination thresholds; to take the hypothetical discrimination threshold corresponding to the maximum Youden index of each syndrome type as the candidate discrimination threshold; and to take the value with the highest repetition from the candidate discrimination thresholds as the discrimination threshold of each syndrome type. The weight coefficient generation module is used to set the total score of each syndrome type to M, calculate the weight coefficient of the syndrome item corresponding to each syndrome type and the adjustment coefficient F of the original weight coefficient using formula 1; and obtain the weight coefficient of the syndrome item corresponding to each syndrome type using formula 2; the maximum hypothesis discrimination threshold + 3 ≥ M ≥ the discrimination threshold; Formula 1: Sum of the original weight coefficients of the syndrome items corresponding to the syndrome type × F = M Formula 2: The weight coefficient of the syndrome item corresponding to the syndrome type is equal to the original weight coefficient of the syndrome item corresponding to the syndrome type × F.

7. A TCM syndrome differentiation and classification system for breast cancer chemotherapy, characterized in that, It contains the TCM syndrome differentiation and classification model for breast cancer chemotherapy as described in claim 1.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the construction method as described in any one of claims 2 to 5.

Images