Method and system for comprehensive evaluation of anti-respiratory virus traditional Chinese medicine in clinic

Abstract

The application discloses a kind of anti respiratory virus Chinese medicine clinical comprehensive evaluation method and system, the method includes: obtaining relevant randomized controlled trial literature, real case data and pharmacoeconomic data;Using Meta analysis method, construct evidence network to obtain multidimensional relative efficacy evaluation results;Relative efficacy evaluation results and pharmacoeconomic data are input into decision tree model, and economic evaluation result is obtained;Real clinical case data, relative efficacy evaluation results and economic evaluation results are combined after division data set;Training set is input into deep confidence neural network model and is trained, and is verified using verification set;Test set is input into optimal deep confidence neural network model, and efficacy prediction result is output;According to efficacy, economic, individualized prediction result calculates comprehensive value score, realizes the comprehensive evaluation of Chinese medicine.The application realizes the accurate evaluation of Chinese medicine efficacy, and provides reliable basis for clinical rational drug use.

Description

Technical Field

[0001] This invention relates to the field of pharmaceutical clinical evaluation technology, specifically to a comprehensive clinical evaluation method and system for traditional Chinese medicine for treating respiratory viruses. Background Technology

[0002] Respiratory viral infections are a major cause of morbidity and mortality worldwide, ranking first among infectious diseases in terms of mortality rate, placing continuous and enormous pressure on public health systems. Traditional Chinese medicine (TCM), with its unique advantages of multi-component, multi-target, and multi-pathway synergistic antiviral action, is widely used in the clinical treatment of respiratory viral infections in my country. TCM preparations have demonstrated significant positive effects in alleviating patients' clinical symptoms, reducing the dosage of hormones, effectively controlling disease progression, and lowering the incidence of complications.

[0003] Traditional Chinese medicine (TCM) preparations are convenient to obtain, easy to carry, and simple to use. During infectious disease pandemics, if mild cases are treated appropriately, they can significantly alleviate the strain on medical resources and reduce ineffective medical costs. However, the core of TCM clinical application lies in syndrome differentiation and individualized medication. Currently, there are numerous types of TCM preparations available in clinical practice, with significant differences in their formulations, yet their indications and functions listed in their instructions are highly similar. This directly leads to frequent instances of indiscriminate use of TCM preparations in clinical practice, severely reducing the therapeutic effect and potentially causing unnecessary adverse reactions.

[0004] Meanwhile, although scientific research and evidence-based medicine regarding the safety and efficacy evaluation of traditional Chinese medicine (TCM) are constantly being updated, most related basic and clinical studies currently suffer from low research quality. There is also a lack of systematic, comprehensive studies on the effectiveness, safety, and pharmacoeconomics of TCM in efficacy evaluations. This results in a lack of high-quality evidence when incorporating TCM into recommended treatment plans, a critical tool for clinical decision-making. This situation not only restricts the rational use of TCM in the treatment of respiratory viral infections but also seriously affects its clinical acceptance and the long-term healthy development of the industry. Summary of the Invention

[0005] In order to solve the above-mentioned technical problems, this application proposes the following technical solution: In a first aspect, embodiments of this application provide a comprehensive clinical evaluation method for traditional Chinese medicines used to treat respiratory viruses, including: Obtain literature on randomized controlled trials, real clinical case data, and publicly available pharmacoeconomic data of traditional Chinese medicine for the treatment of respiratory infections; Based on the aforementioned randomized controlled trial literature, a Bayesian network meta-analysis method was used to construct an evidence network with Western medicine treatment as the common basis for comparison. The effect size and confidence interval of each Chinese patent medicine combined with Western medicine treatment regimen relative to Western medicine treatment alone were calculated. The efficacy of different Chinese patent medicine treatment regimens on each outcome index was ranked by the area under the cumulative ranking probability curve, and multi-dimensional relative efficacy evaluation results were obtained. The relative efficacy evaluation results and the publicly available pharmacoeconomic data are input into a pre-constructed decision tree model. The economics of different traditional Chinese medicine treatment regimens are compared and analyzed using the decision tree model to obtain the economic evaluation results. The real clinical case data, relative efficacy evaluation results and economic evaluation results data are merged and then divided into training set, test set and validation set according to proportion; The training set is input into a pre-built deep belief neural network model for training, a loss function is used for joint optimization, and the optimal deep belief neural network model is selected using the validation set. The test set is input into the optimal deep belief neural network model, and the individualized efficacy prediction results are output. The comprehensive value score of each traditional Chinese medicine (TCM) regimen is calculated based on the relative efficacy evaluation results, economic evaluation results, and individualized efficacy prediction results. The TCM regimens for combating respiratory viruses are evaluated using the comprehensive evaluation score.

[0006] In one possible implementation, a system retrieval strategy is constructed based on the PICOS principle to retrieve literature on randomized controlled trials of traditional Chinese medicine for treating respiratory infections from authoritative domestic and international databases. After deduplication, initial screening, and full-text review of the retrieved literature, the Cochrane risk of bias assessment tool was used to evaluate the quality of the included literature. The calculation formula is as follows: in, The overall score is the result of the bias risk assessment. For document numbering, These are the serial numbers for the risk of bias assessment dimensions. For the first Weight coefficients for each dimension For the first The article is in the first Scoring across multiple dimensions As an indicator variable for dimensional validity, if the first... The dimension in the first A value of 1 indicates that the document was reported in the literature; otherwise, a value of 0 indicates that the document was reported in the literature After selecting qualified literature based on the comprehensive score of the bias risk assessment results, standardized data extraction tables were used to extract literature information. Simultaneously, clinical case data of patients diagnosed with respiratory viral infections and treated with traditional Chinese medicine for respiratory viruses were collected from the information systems, laboratory information systems, and electronic medical record systems of multiple tertiary-level hospitals. Missing value processing, consistency checks, and outlier detection were performed on the collected clinical case data to form structured, real clinical case data. The calculation formulas are as follows: in, For the first The first patient The interpolated value of each indicator, For the first The overall mean of each indicator, For the first The estimated regression coefficients of each indicator with other relevant indicators. For the first A vector of relevant indicators for each patient For the first The mean of relevant indicators for each patient, The random error term follows a normal distribution. , For the first The variance of each indicator, Kappa coefficient This represents the observed consistency rate. For the number of indicator categories, For the first The marginal probability of a row. for Marginal probability of the column For the first The first patient Each indicator value, For the first The first patient The original values ​​of each indicator For the first The average of the indicators, For the first Standard deviation of each indicator; We obtained relevant pharmacoeconomic data on traditional Chinese medicine for treating respiratory viruses from the national drug price database, medical insurance payment database, medical economics research literature, and drug bidding and procurement platforms, and then organized the obtained pharmacoeconomic data to form standardized publicly available pharmacoeconomic data.

[0007] In one possible implementation, based on the aforementioned randomized controlled trial literature, a Bayesian network meta-analysis method is used to construct an evidence network with Western medicine treatment as the common basis for comparison. The effect size and confidence interval of each traditional Chinese medicine combined with Western medicine treatment regimen relative to Western medicine treatment alone are calculated. The efficacy of different traditional Chinese medicine treatment regimens on various outcome indicators is ranked using the area under the cumulative ranking probability curve, resulting in a multi-dimensional relative efficacy evaluation, including: Based on the included randomized controlled trial literature, a network evidence graph was constructed with each traditional Chinese medicine combined with Western medicine treatment regimen as nodes and the direct comparison relationships in the randomized controlled trial literature as edges, including Western medicine treatment alone as a common control. Select the corresponding effect size index according to the type of outcome index, and use the Markov chain Monte Carlo algorithm for Bayesian inference. Based on the inferred effective results, the effect size and confidence interval of each traditional Chinese medicine combined with Western medicine treatment regimen relative to Western medicine treatment alone were calculated. Simultaneously, the area under the cumulative ranking probability curve for each combination of traditional Chinese medicine and Western medicine treatment regimen was calculated for each outcome indicator. The calculation formula is as follows: in, For the first A treatment plan combining traditional Chinese medicine and Western medicine was implemented in the first phase. The area under the cumulative probability ranking curve for each outcome metric. To participate in the The total number of treatment regimens evaluated by each outcome indicator. For the first A treatment plan combining traditional Chinese medicine and Western medicine was implemented in the first phase. Ranking on each outcome metric For the first A treatment plan combining traditional Chinese medicine and Western medicine was implemented in the first phase. The probability that the ranking on any outcome indicator does not exceed the mth position; The efficacy of different traditional Chinese medicine treatment regimens was ranked across various outcome indicators by the area under the cumulative probability curve, resulting in a multi-dimensional relative efficacy evaluation.

[0008] In one possible implementation, the corresponding effect size index is selected based on the type of outcome index, and Bayesian inference is performed using the Markov chain Monte Carlo algorithm. Based on the valid inference results, the effect size and confidence interval of each traditional Chinese medicine combined with Western medicine treatment regimen relative to Western medicine treatment alone are calculated, including: For binary outcome indicators, the odds ratio was used as the effect size. The effect size and confidence interval of each traditional Chinese medicine combined with Western medicine treatment regimen relative to Western medicine alone were calculated for each binary outcome indicator. The calculation formula is as follows: in, For the first The ratio of traditional Chinese medicine combined with Western medicine treatment to Western medicine alone. For the first The number of cases of effectiveness or adverse reactions of the combination therapy regimen. For the first The total sample size of the combination therapy regimens This refers to the number of cases where Western medicine alone provides effective treatment or causes adverse reactions. This represents the total sample size for Western medicine-only treatments. for The natural logarithm of follows a normal distribution. The standard normal distribution quantiles corresponding to the 95% confidence level. for The 95% confidence interval; For continuous variable outcome indicators, the mean difference was used as the effect size. The effect size and confidence interval of each traditional Chinese medicine combined with Western medicine treatment regimen relative to Western medicine alone were calculated for each continuous variable outcome indicator. The calculation formula is as follows: in, For the first The mean difference in target continuity outcome indicators between the combined Chinese patent medicine and Western medicine treatment regimens and Western medicine treatment regimens alone. For the first The sample mean of the target continuous outcome indicators in the combination therapy regimens. This represents the sample mean of the target continuous outcome indicators in Western medicine-only treatments. For the first The sample standard deviation of the target continuity outcome measure in various combination therapy regimens. This represents the sample standard deviation of the target continuous outcome indicator in Western medicine-only treatment. for The 95% confidence interval.

[0009] In one possible implementation, the relative efficacy evaluation results and the publicly available pharmacoeconomic data are input into a pre-constructed decision tree model. The decision tree model is then used to perform a comparative economic analysis of different traditional Chinese medicine treatment regimens to obtain economic evaluation results, including: After extracting the effect size and confidence interval of each traditional Chinese medicine (TCM) combined with Western medicine treatment regimen relative to Western medicine alone under the overall clinical efficacy outcome index, the effect size of each TCM combined with Western medicine treatment regimen relative to Western medicine treatment is converted into the absolute efficacy in the decision tree model. The calculation formula is as follows: in, For the first The absolute effectiveness of a combination of traditional Chinese medicine and Western medicine treatment in a decision tree model. The overall clinical efficacy rate was the highest among the clinical outcome indicators. The ratio of traditional Chinese medicine combined with Western medicine treatment to Western medicine alone. The baseline efficacy rate for Western medicine treatment alone. Subgroup correction factor To publish the bias correction factor, The total number of subgroups, For the first Sample size weights for each subgroup For the first The treatment in the first Efficiency in each subgroup For the first The overall effectiveness rate of this treatment To determine the number of studies included, For the first The first treatment The logarithmic effect size of each study For the first The mean of the logarithmic effect sizes from all studies on this treatment. For the first Standard error of the therapeutic effect size; The total cost of different combined traditional Chinese medicine and Western medicine treatment regimens is calculated according to cost accounting rules. The calculation formula is as follows: in, For the first The total cost of a treatment plan combining traditional Chinese medicine and Western medicine. The average daily hospitalization cost, For the first The average daily cost of using this type of traditional Chinese medicine. For the first The average daily cost of auxiliary drugs required for various traditional Chinese medicine preparations For the first The average length of hospital stay for each treatment option This represents the total number of adverse reaction types. For the first The first treatment regimen occurred The probability of adverse reactions. To process the first The average daily drug cost for each type of adverse reaction. For the reason The average daily cost of additional hospitalization due to adverse reactions. For the reason The average length of hospital stay due to adverse reactions This represents the total number of laboratory monitoring items. For the first This type of treatment requires additional monitoring. The frequency of each project For the first The cost of a single monitoring project The average length of hospital stay for Western medicine treatment alone; The absolute efficiency and the calculated total cost are input into the decision tree model for cost-effectiveness analysis, and the results are sorted from low to high cost. The calculation formula is as follows: , ; in, The expected efficacy values ​​for each treatment regimen are given. Cost-effectiveness ratio of each treatment option Weighting of life years for clinically effective status based on quality. Weighting of quality-adjusted life years for clinically ineffective states; After eliminating the absolutely disadvantageous solutions with high costs but low outputs, calculate the incremental cost-effectiveness ratio for each pair of the remaining solutions. The calculated incremental cost-effectiveness ratio is compared with the preset willingness-to-pay threshold. If the incremental cost-effectiveness ratio is lower than the preset willingness-to-pay threshold, it indicates that the incremental cost is economical and a better treatment option should be recommended. If the incremental cost-effectiveness ratio is higher than the preset willingness-to-pay threshold, the lower-cost option will be selected first.

[0010] In one possible implementation, after eliminating the absolutely inferior solutions that are high-cost but low-output, the formula for calculating the incremental cost-effectiveness ratio for pairwise comparisons of the remaining solutions is as follows: in, The incremental cost-effectiveness ratio for pairwise comparison of the remaining options. and The first The cost and effectiveness of this treatment plan and The costs and effects of the reference scheme are respectively. and These represent the variance of the cost estimate. and These represent the variance of the effect estimate.

[0011] In one possible implementation, the training set is input into a pre-built deep belief neural network model for training, a loss function is used for joint optimization, and a validation set is used to select the optimal deep belief neural network model, including: An unsupervised greedy layer-by-layer training method is adopted to pre-train each hidden layer of the deep belief network. Based on the input layer data, the feature representation of each layer is learned by a restricted Boltzmann machine (RBM). First, train the RBM consisting of the input layer and the first hidden layer. Then, use the output of the first hidden layer as the input of the second RBM for training. Continue this process to complete the pre-training of all hidden layers and initialize the model weight parameters. The pre-trained model weights are used as initial parameters and connected to the output layer to form a complete deep belief neural network. A supervised learning approach is adopted to match the input features of the training set with the corresponding real clinical outcome labels. The weights and biases of the entire network are adjusted through the backpropagation algorithm. After joint optimization using a loss function, the optimal deep belief neural network model is selected using a validation set.

[0012] In one possible implementation, the loss function is calculated as follows: in, For loss function, The total number of samples in the training set. For sample index, For category indexing, For the first The sample at the th Real labels in each category The output of the Softmax layer The sample at the th Predict probabilities for each category. The L2 regularization coefficient is... The squared L2 norm of all weight parameters, This is the sparsity penalty coefficient. For hidden layer index, The total number of hidden layers in a deep confidence neural network model. KL divergence is used to constrain the average activation rate of neurons in the hidden layer. Approximate target sparsity parameter .

[0013] In one possible implementation, the formula for calculating the comprehensive value score of each traditional Chinese medicine regimen based on the relative efficacy evaluation results, economic evaluation results, and individualized efficacy prediction results is as follows: in, The overall value score for each traditional Chinese medicine prescription is calculated. The relative weights of the efficacy evaluation dimensions As the weight of the economic evaluation dimension, Weights for individualized efficacy prediction dimensions This refers to the total number of outcome indicators used in the efficacy evaluation. For the first A treatment plan combining traditional Chinese medicine and Western medicine was implemented in the first phase. The area under the cumulative probability ranking curve for each outcome metric. For the first Incremental cost-effectiveness ratio of the proposed solution to the reference solution; and These represent the maximum and minimum ICER values ​​for all included evaluation schemes. Predicting the first [number]th ... The efficacy accuracy of this treatment in the target patient population. and These represent the maximum and minimum accuracy rates of individualized efficacy prediction across all included evaluation protocols.

[0014] Secondly, embodiments of this application provide a comprehensive clinical evaluation system for traditional Chinese medicine for treating respiratory viruses, including: The acquisition module is used to acquire literature on randomized controlled trials of traditional Chinese medicine for treating respiratory infections, real clinical case data, and publicly available pharmacoeconomic data. The relative efficacy evaluation module is used to construct an evidence network with Western medicine treatment as the common comparison basis based on the randomized controlled trial literature using the Bayesian network meta-analysis method. It calculates the effect size and confidence interval of each Chinese patent medicine combined with Western medicine treatment regimen relative to Western medicine treatment alone, and ranks the efficacy of different Chinese patent medicine treatment regimens on each outcome index by the area under the cumulative ranking probability curve, so as to obtain multi-dimensional relative efficacy evaluation results. The economic comparison analysis module is used to input the relative efficacy evaluation results and the publicly available pharmacoeconomic data into a pre-constructed decision tree model, and to conduct an economic comparison analysis of different traditional Chinese medicine treatment plans through the decision tree model to obtain economic evaluation results. The merging and partitioning module is used to merge the real clinical case data, relative efficacy evaluation results, and economic evaluation results data and then divide them into training set, test set, and validation set according to a certain ratio. The training module is used to input the training set into a pre-built deep belief neural network model for training, perform joint optimization using a loss function, and select the optimal deep belief neural network model using a validation set. The efficacy prediction module is used to input the test set into the optimal deep belief neural network model and output individualized efficacy prediction results. The comprehensive evaluation module is used to calculate the comprehensive value score of each traditional Chinese medicine prescription based on the relative efficacy evaluation results, economic evaluation results, and individualized efficacy prediction results, and to evaluate the traditional Chinese medicine prescriptions against respiratory viruses through the comprehensive evaluation score.

[0015] Compared with the prior art, the beneficial effects of this application are as follows: This application overcomes the limitations of a single data source by integrating three core data sources: randomized controlled trial literature, real clinical case data, and publicly available pharmacoeconomic data. At the same time, it combines data analysis methods such as Bayesian network meta-analysis, decision tree models, and deep belief neural network models to achieve a comprehensive evaluation of traditional Chinese medicine preparations from efficacy and cost-effectiveness to individualized efficacy prediction, providing scientific and rigorous evidence support for the clinical evaluation of traditional Chinese medicine preparations.

[0016] This application constructs an evidence network based on Western medicine monotherapy as a common comparative basis, selects appropriate effect size indicators for different types of outcome indicators, and achieves accurate efficacy ranking of various traditional Chinese medicine combination therapy regimens on different outcome indicators by using the area under the cumulative ranking probability curve. It clearly presents the efficacy advantages and applicable scenarios of different traditional Chinese medicines, making up for the shortcomings of traditional efficacy evaluation being vague and lacking quantitative standards.

[0017] This application utilizes a deep belief neural network model, integrating real clinical case characteristics, efficacy, and cost-effectiveness data for model training and prediction. It can output individualized efficacy prediction results for different patients, enabling the clinical selection of traditional Chinese medicine to shift from experience-based to precision-based, which aligns with the core essence of TCM syndrome differentiation and treatment and individualized drug administration, effectively improving the rationality and effectiveness of TCM clinical drug use. Attached Figure Description

[0018] Figure 1 A flowchart illustrating a comprehensive clinical evaluation method for traditional Chinese medicine for treating respiratory viruses, provided in this application embodiment; Figure 2 A flowchart of the document screening process provided for embodiments of this application; Figure 3 A network diagram of the various outcome indicators provided in the embodiments of this application; Figure 4 The cumulative probability ranking curves of the various indicator outcomes provided in the embodiments of this application are sorted by area under the curve; Figure 5 A decision tree model for treating viral pneumonia with different traditional Chinese medicine injections provided in the embodiments of this application; Figure 6 The result diagram of the decision tree model for treating viral pneumonia with different traditional Chinese medicine injections provided in the embodiments of this application; Figure 7 The deep belief neural network model provided in the embodiments of this application. Detailed Implementation

[0019] The present solution will now be described in conjunction with the accompanying drawings and specific embodiments.

[0020] Figure 1 A flowchart illustrating a comprehensive clinical evaluation method for traditional Chinese medicine for treating respiratory viruses, provided in this application embodiment, is shown below. Figure 1 This embodiment includes a comprehensive clinical evaluation method for a traditional Chinese medicine for treating respiratory viruses, comprising: S101, obtain literature on randomized controlled trials, real clinical case data and publicly available pharmacoeconomic data of traditional Chinese medicine for the treatment of respiratory tract infections.

[0021] See Figure 2 In this embodiment, a system retrieval strategy is constructed based on the PICOS principle. Randomized controlled trials of traditional Chinese medicine for treating respiratory infections using antiviral drugs are retrieved from authoritative domestic and international databases. Evidence is primarily retrieved from Cochrane, PubMed, Medline, Embase, UpToDate, Clinical Trials, CNKI, Wanfang Database, and VIP Database, covering the past 10 years. The retrieval strategy combines subject terms and free terms. After deduplication, initial screening, and full-text review of the retrieved literature, the Cochrane risk of bias assessment tool is used to evaluate the quality of the included literature. The calculation formula is as follows: in, The overall score is the result of the bias risk assessment. For document numbering, These are the serial numbers for the risk of bias assessment dimensions. For the first Weight coefficients for each dimension For the first The article is in the first Scoring across multiple dimensions As an indicator variable for dimensional validity, if the first... The dimension in the first A score of 1 indicates that the article was reported in the literature; otherwise, a score of 0 indicates that it was reported in the literature. After screening qualified literature based on the comprehensive score of the risk of bias assessment results, standardized data extraction tables were used to extract literature information. The extracted content mainly included: basic research information, including the first author and publication year; characteristics of the research subjects, including sample size, gender composition, mean age or age range, intervention measures, treatment duration, outcome indicators and data results, research design type, and main factors for risk of bias assessment.

[0022] In this embodiment, a total of 74 RCTs were included, with a sample size of 7463 cases, including 3739 cases in the experimental group and 3724 cases in the control group. The intervention measures in the control group were all conventional Western medicine treatment alone, while the intervention measures in the experimental group were 8 kinds of traditional Chinese medicine injections combined with conventional Western medicine. The 8 kinds of traditional Chinese medicine injections involved were: Retoxin Injection (RDN) 22 articles, Xiyanping Injection (XYP) 22 articles, Tanreqing Injection (TRQ) 16 articles, Qingkailing Injection (QKL) 3 articles, Yanhuning Injection (YHN) 8 articles, Ezhu Oil Injection (EZY) 1 article, Xuebijing Injection (XBJ) 1 article, and Shenqi Fuzheng Injection (SQFZ) 1 article.

[0023] Simultaneously, clinical case data of patients diagnosed with respiratory viral infections and treated with traditional Chinese medicine for respiratory viruses were collected from the information systems, laboratory information systems, and electronic medical record systems of multiple tertiary-level hospitals. This included patients' demographic characteristics, diagnostic information, clinical symptoms and signs, laboratory test indicators, imaging examinations, treatment plans, and clinical outcomes. Missing value processing, consistency checks, and outlier detection were performed on the collected clinical case data to form structured, real-world clinical case data. The calculation formulas are as follows: in, For the first The first patient The interpolated value of each indicator, For the first The overall mean of each indicator, For the first The estimated regression coefficients of each indicator with other relevant indicators. For the first A vector of relevant indicators for each patient For the first The mean of relevant indicators for each patient, The random error term follows a normal distribution. , For the first The variance of each indicator, Kappa coefficient This represents the observed consistency rate. For the number of indicator categories, For the first The marginal probability of a row. for Marginal probability of the column For the first The first patient Each indicator value, For the first The first patient The original values ​​of each indicator For the first The average of the indicators, For the first Standard deviation of each indicator; We obtained relevant pharmacoeconomic data on traditional Chinese medicine for treating respiratory viruses from the national drug price database, medical insurance payment database, medical economics research literature, and drug bidding and procurement platforms, and then organized the obtained pharmacoeconomic data to form standardized publicly available pharmacoeconomic data.

[0024] We obtained data on average daily hospitalization costs for pneumonia, average daily bed fees, various examination and testing fees, and adverse reaction management fees from the China Health Statistics Yearbook, the China Price Statistics Yearbook, and medical service price manuals of various provinces and cities. We also obtained the quality-adjusted life years weights for respiratory virus infection-related health statuses from published pharmacoeconomic literature or the EQ-5D and SF-6D health utility value databases.

[0025] S102. Based on the literature of randomized controlled trials, a Bayesian network meta-analysis method was used to construct an evidence network with Western medicine treatment as the common basis for comparison. The effect size and confidence interval of each Chinese patent medicine combined with Western medicine treatment regimen relative to Western medicine treatment alone were calculated. The efficacy of different Chinese patent medicine treatment regimens on each outcome index was ranked by the area under the cumulative ranking probability curve, and multi-dimensional relative efficacy evaluation results were obtained.

[0026] See Figure 3In this embodiment, based on the included randomized controlled trial literature, a network evidence graph is constructed, using each traditional Chinese medicine combined with Western medicine treatment regimen as nodes and the direct comparison relationships in the randomized controlled trial literature as edges. This network includes Western medicine treatment alone as a common control. Each dot represents an intervention, and the size of the dot is proportional to the sample size included in that intervention. The connecting lines represent the comparison relationships between different interventions, and their thickness is related to the number of included studies. In the graph, A represents the overall clinical effectiveness rate, B represents the cough relief time, C represents the rales relief time, D represents the fever reduction time, E represents the hospital stay, F represents the CRA outcome, G represents the IL-6 outcome, and H represents the TNF-α outcome. Figure 4 The same applies to China.

[0027] Based on the type of outcome indicator, the corresponding effect size indicator is selected, and Bayesian inference is performed using the Markov chain Monte Carlo algorithm. Based on the effective inference results, the effect size and confidence interval of each traditional Chinese medicine combined with Western medicine treatment regimen relative to Western medicine treatment alone are calculated. In this embodiment, the outcome indicators include binary indicators (overall clinical effectiveness rate, incidence of adverse reactions) and continuous indicators (cough relief time, time to disappearance of lung rales, fever reduction time, length of hospital stay, C-reactive protein (CRP), interleukin-6 (IL-6), and tumor necrosis factor-α (TNF-α).

[0028] For binary outcome indicators, the odds ratio was used as the effect size. The effect size and confidence interval of each traditional Chinese medicine combined with Western medicine treatment regimen relative to Western medicine alone were calculated for each binary outcome indicator. The calculation formula is as follows: in, For the first The ratio of traditional Chinese medicine combined with Western medicine treatment to Western medicine alone. For the first The number of cases of effectiveness or adverse reactions of the combination therapy regimen. For the first The total sample size of the combination therapy regimens This refers to the number of cases where Western medicine alone provides effective treatment or causes adverse reactions. This represents the total sample size for Western medicine-only treatments. for The natural logarithm of follows a normal distribution. The standard normal distribution quantiles corresponding to the 95% confidence level. for The 95% confidence interval.

[0029] For continuous variable outcome indicators, the mean difference was used as the effect size. The effect size and confidence interval of each traditional Chinese medicine combined with Western medicine treatment regimen relative to Western medicine alone were calculated for each continuous variable outcome indicator. The calculation formula is as follows: in, For the first The mean difference in target continuity outcome indicators between the combined Chinese patent medicine and Western medicine treatment regimens and Western medicine treatment regimens alone. For the first The sample mean of the target continuous outcome indicators in the combination therapy regimens. This represents the sample mean of the target continuous outcome indicators in Western medicine-only treatments. For the first The sample standard deviation of the target continuity outcome measure in various combination therapy regimens. This represents the sample standard deviation of the target continuous outcome indicator in Western medicine-only treatment. for The 95% confidence interval.

[0030] Simultaneously, the area under the cumulative ranking probability curve for each combination of traditional Chinese medicine and Western medicine treatment regimen was calculated for each outcome indicator. The calculation formula is as follows: in, For the first A treatment plan combining traditional Chinese medicine and Western medicine was implemented in the first phase. The area under the cumulative probability ranking curve for each outcome metric. To participate in the The total number of treatment regimens evaluated by each outcome indicator. For the first A treatment plan combining traditional Chinese medicine and Western medicine was implemented in the first phase. Ranking on each outcome metric For the first A treatment plan combining traditional Chinese medicine and Western medicine was implemented in the first phase. The probability that the ranking on any outcome indicator does not exceed the mth position; The efficacy of different traditional Chinese medicine treatment regimens across various outcome indicators was ranked using the area under the cumulative probability curve, yielding a multi-dimensional relative efficacy evaluation. The ranking results are as follows: Figure 4As shown in the figure, current evaluation results indicate that Yanhuning injection performs exceptionally well across multiple indicators, being the optimal choice for shortening cough relief time (SUCRA=78.1%), lung rales relief time (SUCRA=76.0%), hospital stay (SUCRA=82.8%), and improving CRP (SUCRA=76.9%) and TNF-α (SUCRA=95.8%). Qingkailing injection demonstrates significant advantages in improving overall clinical efficacy (SUCRA=75.5%) and reducing fever (SUCRA=82.7%). Tanreqing injection is the best intervention for improving IL-6 levels (SUCRA=82.2%). Shenqi Fuzheng injection ranks first in shortening fever reduction time (SUCRA=89.7%). Overall, the SUCRA values ​​of all combined TCM injection regimens are significantly higher than those of conventional Western medicine alone.

[0031] S103. The relative efficacy evaluation results and publicly available pharmacoeconomic data are input into a pre-constructed decision tree model. The economic comparative analysis of different traditional Chinese medicine treatment plans is conducted through the decision tree model to obtain the economic evaluation results.

[0032] See Figure 5 In this embodiment, the decision tree model simulates the decision-making process through a tree structure. The model begins with a decision node (□), leading to nine decision branches, representing conventional treatment and eight combined treatment regimens of traditional Chinese medicine injections. Each decision branch is connected to an opportunity node (○), which branches into two opportunity branches, representing two possible outcomes of the regimen: clinically effective and clinically ineffective. The endpoint of each path is a terminal node (△), which is linked to the cumulative cost and health output of that path. The parameters and definitions in the decision tree model are shown in Table 1. Table 1. Decision Tree Model Parameters and Definitions In this embodiment, after extracting the effect size and confidence interval of each traditional Chinese medicine combined with Western medicine treatment regimen relative to Western medicine alone under the overall clinical effectiveness outcome index, the effect size of each traditional Chinese medicine combined with Western medicine treatment regimen relative to Western medicine treatment is converted into the absolute effectiveness in the decision tree model. The calculation formula is as follows: in, For the first The absolute effectiveness of a combination of traditional Chinese medicine and Western medicine treatment in a decision tree model. The overall clinical efficacy rate was the highest among the clinical outcome indicators. The ratio of traditional Chinese medicine combined with Western medicine treatment to Western medicine alone. The baseline efficacy rate for Western medicine treatment alone. Subgroup correction factor To publish the bias correction factor, The total number of subgroups, For the first Sample size weights for each subgroup For the first The treatment in the first Efficiency in each subgroup For the first The overall effectiveness rate of this treatment To determine the number of studies included, For the first The first treatment The logarithmic effect size of each study For the first The mean of the logarithmic effect sizes from all studies on this treatment. For the first Standard error of the therapeutic effect size; The total cost of different combined traditional Chinese medicine and Western medicine treatment regimens is calculated according to cost accounting rules. The calculation formula is as follows: in, For the first The total cost of a treatment plan combining traditional Chinese medicine and Western medicine. The average daily hospitalization cost, For the first The average daily cost of using this type of traditional Chinese medicine. For the first The average daily cost of auxiliary drugs required for various traditional Chinese medicine preparations For the first The average length of hospital stay for each treatment option This represents the total number of adverse reaction types. For the first The first treatment regimen occurred The probability of adverse reactions. To process the first The average daily drug cost for each type of adverse reaction. For the reason The average daily cost of additional hospitalization due to adverse reactions. For the reason The average length of hospital stay due to adverse reactions This represents the total number of laboratory monitoring items. For the first This type of treatment requires additional monitoring. The frequency of each project For the first The cost of a single monitoring project The average length of hospital stay for Western medicine treatment alone; The absolute efficiency and the total calculated cost are input into the decision tree model for cost-effectiveness analysis, and the results are sorted from low to high cost. The calculation formula is as follows: , ; in, The expected efficacy values ​​for each treatment regimen are given. Cost-effectiveness ratio of each treatment option Weighting of life years for clinically effective status based on quality. Weighting of quality-adjusted life years for clinically ineffective states.

[0033] The cost-effectiveness analysis results are shown in Table 2: Table 2. Cost-effectiveness analysis of different traditional Chinese medicines combined with Western medicine treatments Analysis shows that compared with conventional Western medicine treatment (CT) alone, the combined treatment regimens of the eight traditional Chinese medicine injections all exhibited significant cost savings and efficacy advantages, demonstrating superior economic efficiency within the framework of health economics. Among them, the combined treatment with Yanhuning had the lowest total cost, saving 1956.81 yuan compared to CT. The combined treatment with Qingkailing achieved the best therapeutic effect, with a total clinical effective rate of 95.34%, an improvement of 17% compared to CT. Both Yanhuning and Qingkailing regimens showed absolute advantages in direct comparison with conventional treatments, namely lower cost and better efficacy.

[0034] To further differentiate the economics of Qingkailing and Yanhuning, an incremental cost-effectiveness analysis is conducted. After eliminating the absolutely disadvantageous option with high cost but low output, the incremental cost-effectiveness ratio of the remaining options is calculated pairwise. The formula is as follows: in, The incremental cost-effectiveness ratio for pairwise comparison of the remaining options. and The first The cost and effectiveness of this treatment plan and The costs and effects of the reference scheme are respectively. and These represent the variance of the cost estimate. and These represent the variance of the effect estimate. The results are shown in Table 3: Table 3. Incremental cost-effectiveness analysis of different traditional Chinese medicines combined with Western medicine treatments. The results showed that, compared with the lowest-cost Yanhuning combination therapy, the ICER of the Qingkailing combination therapy was 1322.68 yuan. This ICER value is far lower than the WTP, indicating that the incremental cost paid for the Qingkailing regimen to obtain additional health benefits is significantly economical. Therefore, the Qingkailing combination therapy can be regarded as the preferred option with the greatest cost-effectiveness, while the Yanhuning regimen is the lowest-cost and most effective alternative.

[0035] The calculated incremental cost-effectiveness ratio is compared with a preset willingness-to-pay threshold. If the incremental cost-effectiveness ratio is lower than the preset willingness-to-pay threshold, it indicates that the incremental cost is economical, and a better treatment option should be recommended. If the incremental cost-effectiveness ratio is higher than the preset willingness-to-pay threshold, the lower-cost option is preferred. The decision tree model results are as follows: Figure 6 As shown, the results indicate that the combination of Yanhuning injection and conventional Western medicine treatment group has lower costs and better efficacy, and the decision tree determines that it is more cost-effective.

[0036] S104, after merging real clinical case data, relative efficacy evaluation results, and cost-effectiveness evaluation results, divides them proportionally into training set, test set, and validation set.

[0037] In this embodiment, eight key clinical features are extracted from real clinical case data as model inputs, and relative efficacy evaluation results and cost-effectiveness evaluation results are used as supervision labels.

[0038] S105, the training set is input into the pre-built deep belief neural network model for training, the loss function is used for joint optimization, and the optimal deep belief neural network model is selected using the validation set.

[0039] In this embodiment, the deep belief neural network model includes an input layer, which is connected to a first hidden layer, a second hidden layer, a third hidden layer, and an output layer. Each hidden layer contains a residual unit.

[0040] See Figure 7 In this embodiment, the input layer has 8 neurons, corresponding to 8 clinical features; the first hidden layer has 64 neurons; the second hidden layer has 32 neurons; and the third hidden layer has 16 neurons. An unsupervised greedy layer-by-layer training method is used to pre-train each hidden layer of the deep belief network. Based on the input layer data, the feature representation of each layer is learned through a restricted Boltzmann machine (RBM). First, train the RBM consisting of the input layer and the first hidden layer. Then, use the output of the first hidden layer as the input of the second RBM for training. Continue this process to complete the pre-training of all hidden layers and initialize the model weight parameters. The pre-trained model weights are used as initial parameters and connected to the output layer to form a complete deep belief neural network. A supervised learning approach is adopted to match the input features of the training set with the corresponding real clinical outcome labels. The weights and biases of the entire network are adjusted through the backpropagation algorithm. Joint optimization is performed using a loss function, and the optimal deep confidence neural network model is selected using the validation set. The formula for calculating the loss function is as follows: in, For loss function, The total number of samples in the training set. For sample index, For category indexing, For the first The sample at the th Real labels in each category The output of the Softmax layer The sample at the th Predict probabilities for each category. The L2 regularization coefficient is... The squared L2 norm of all weight parameters, This is the sparsity penalty coefficient. For hidden layer index, The total number of hidden layers in a deep confidence neural network model. KL divergence is used to constrain the average activation rate of neurons in the hidden layer. Approximate target sparsity parameter .

[0041] The Adam optimizer was used for parameter updates. The initial learning rate was set to 0.001, the decay factor was 0.95, the batch size was 64, the training epochs were 200, and the early stopping method was used to stop training when the validation set loss no longer decreased for 10 consecutive epochs.

[0042] S106, input the test set into the optimal deep belief neural network model, and output the individualized efficacy prediction results.

[0043] S107. The comprehensive value score of each traditional Chinese medicine regimen is calculated based on the relative efficacy evaluation results, economic evaluation results, and individualized efficacy prediction results. The comprehensive evaluation score is used to evaluate the traditional Chinese medicine regimen against respiratory viruses.

[0044] In this embodiment, the comprehensive value score is calculated using a weighted summation method, integrating the relative efficacy evaluation results, the economic evaluation results, and the individualized efficacy prediction results. The calculation formula is as follows: in, The overall value score for each traditional Chinese medicine prescription is calculated. The relative weights of the efficacy evaluation dimensions As the weight of the economic evaluation dimension, Weights for individualized efficacy prediction dimensions This refers to the total number of outcome indicators used in the efficacy evaluation. For the first A treatment plan combining traditional Chinese medicine and Western medicine was implemented in the first phase. The area under the cumulative probability ranking curve for each outcome metric. For the first Incremental cost-effectiveness ratio of the proposed solution to the reference solution; and These represent the maximum and minimum ICER values ​​for all included evaluation schemes. Predicting the first [number]th ... The efficacy accuracy of this treatment in the target patient population. and These represent the maximum and minimum accuracy rates of individualized efficacy prediction across all included evaluation protocols.

[0045] Corresponding to the clinical comprehensive evaluation method for antiviral traditional Chinese medicine provided in the above embodiments, this application also provides an embodiment of a clinical comprehensive evaluation system for antiviral traditional Chinese medicine.

[0046] This application provides a comprehensive clinical evaluation system for traditional Chinese medicine for treating respiratory viruses, comprising: The acquisition module is used to acquire literature on randomized controlled trials, real clinical case data, and publicly available pharmacoeconomic data of traditional Chinese medicine for the treatment of respiratory infections.

[0047] The relative efficacy evaluation module is used to construct an evidence network based on Western medicine treatment as a common comparison basis, using Bayesian network meta-analysis, according to randomized controlled trial literature. It calculates the effect size and confidence interval of each Chinese patent medicine combined with Western medicine treatment regimen relative to Western medicine treatment alone, and ranks the efficacy of different Chinese patent medicine treatment regimens on each outcome indicator by the area under the cumulative ranking probability curve, thus obtaining multi-dimensional relative efficacy evaluation results.

[0048] The economic comparison analysis module is used to input the relative efficacy evaluation results and the publicly available pharmacoeconomic data into a pre-constructed decision tree model. The decision tree model is used to conduct an economic comparison analysis of different traditional Chinese medicine treatment plans to obtain economic evaluation results.

[0049] The merge and partition module is used to merge real clinical case data, relative efficacy evaluation results, and economic evaluation results data and then divide them into training set, test set, and validation set according to a certain ratio.

[0050] The training module is used to input the training set into a pre-built deep belief neural network model for training, perform joint optimization using a loss function, and select the optimal deep belief neural network model using a validation set.

[0051] The efficacy prediction module is used to input the test set into the optimal deep belief neural network model and output individualized efficacy prediction results.

[0052] The comprehensive evaluation module is used to calculate the comprehensive value score of each traditional Chinese medicine regimen based on the relative efficacy evaluation results, economic evaluation results, and individualized efficacy prediction results. The comprehensive evaluation score is used to evaluate traditional Chinese medicine regimens against respiratory viruses.

[0053] In this embodiment, "multiple" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent the existence of A alone, the simultaneous existence of A and B, or the existence of B alone. A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0054] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0055] The above description is merely a specific embodiment of this application. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the protection scope of this application. The protection scope of this application should be determined by the protection scope of the claims.

Claims

1. A comprehensive clinical evaluation method for traditional Chinese medicine preparations for treating respiratory viruses, characterized in that, include: Obtain literature on randomized controlled trials, real clinical case data, and publicly available pharmacoeconomic data of traditional Chinese medicine for the treatment of respiratory infections; Based on the aforementioned randomized controlled trial literature, a Bayesian network meta-analysis method was used to construct an evidence network with Western medicine treatment as the common basis for comparison. The effect size and confidence interval of each Chinese patent medicine combined with Western medicine treatment regimen relative to Western medicine treatment alone were calculated. The efficacy of different Chinese patent medicine treatment regimens on each outcome index was ranked by the area under the cumulative ranking probability curve, and multi-dimensional relative efficacy evaluation results were obtained. The relative efficacy evaluation results and the publicly available pharmacoeconomic data are input into a pre-constructed decision tree model. The economics of different traditional Chinese medicine treatment regimens are compared and analyzed using the decision tree model to obtain the economic evaluation results. The real clinical case data, relative efficacy evaluation results and economic evaluation results data are merged and then divided into training set, test set and validation set according to proportion; The training set is input into a pre-built deep belief neural network model for training, a loss function is used for joint optimization, and the optimal deep belief neural network model is selected using the validation set. The test set is input into the optimal deep belief neural network model, and the individualized efficacy prediction results are output. The comprehensive value score of each traditional Chinese medicine (TCM) regimen is calculated based on the relative efficacy evaluation results, economic evaluation results, and individualized efficacy prediction results. The TCM regimens for combating respiratory viruses are evaluated using the comprehensive evaluation score.

2. The clinical comprehensive evaluation method for antiviral traditional Chinese medicines according to claim 1, characterized in that, The acquisition of randomized controlled trial literature, real clinical case data, and publicly available pharmacoeconomic data on traditional Chinese medicine for treating respiratory infections includes: Based on the PICOS principle, a systematic search strategy was constructed to retrieve randomized controlled trials of traditional Chinese medicine for treating respiratory infections from authoritative domestic and international databases. After deduplication, initial screening, and full-text review of the retrieved literature, the Cochrane risk of bias assessment tool was used to evaluate the quality of the included literature. The calculation formula is as follows: in, The overall score is the result of the bias risk assessment. For document numbering, These are the serial numbers for the risk of bias assessment dimensions. For the first Weight coefficients for each dimension For the first The article is in the first Scoring across multiple dimensions As an indicator variable for dimensional validity, if the first... The dimension in the first A value of 1 indicates that the document was reported in the literature; otherwise, a value of 0 indicates that the document was reported in the literature. After selecting qualified literature based on the comprehensive score of the bias risk assessment results, standardized data extraction tables were used to extract literature information. Simultaneously, clinical case data of patients diagnosed with respiratory viral infections and treated with traditional Chinese medicine for respiratory viruses were collected from the information systems, laboratory information systems, and electronic medical record systems of multiple tertiary-level hospitals. Missing value processing, consistency checks, and outlier detection were performed on the collected clinical case data to form structured, real clinical case data. The calculation formulas are as follows: in, For the first The first patient The interpolated value of each indicator, For the first The overall mean of each indicator, For the first The estimated regression coefficients of each indicator with other relevant indicators. For the first A vector of relevant indicators for each patient For the first The mean of relevant indicators for each patient, The random error term follows a normal distribution. , For the first The variance of each indicator, Kappa coefficient This represents the observed consistency rate. For the number of indicator categories, For the first The marginal probability of a row. for Marginal probability of the column For the first The first patient Each indicator value, For the first The first patient The original values ​​of each indicator For the first The average of the indicators, For the first Standard deviation of each indicator; We obtained relevant pharmacoeconomic data on traditional Chinese medicine for treating respiratory viruses from the national drug price database, medical insurance payment database, medical economics research literature, and drug bidding and procurement platforms, and then organized the obtained pharmacoeconomic data to form standardized publicly available pharmacoeconomic data.

3. The clinical comprehensive evaluation method for antiviral traditional Chinese medicines according to claim 1, characterized in that, Based on the aforementioned randomized controlled trial literature, a Bayesian network meta-analysis was used to construct an evidence network with Western medicine treatment as the common basis for comparison. The effect sizes and confidence intervals of each traditional Chinese medicine (TCM) combined with Western medicine treatment regimens relative to Western medicine treatment alone were calculated. The efficacy of different TCM treatment regimens across various outcome indicators was ranked using the area under the cumulative probability curve, resulting in a multi-dimensional relative efficacy evaluation, including: Based on the included randomized controlled trial literature, a network evidence graph was constructed with each traditional Chinese medicine combined with Western medicine treatment regimen as nodes and the direct comparison relationships in the randomized controlled trial literature as edges, including Western medicine treatment alone as a common control. Select the corresponding effect size index according to the type of outcome index, and use the Markov chain Monte Carlo algorithm for Bayesian inference. Based on the inferred effective results, the effect size and confidence interval of each traditional Chinese medicine combined with Western medicine treatment regimen relative to Western medicine treatment alone were calculated. Simultaneously, the area under the cumulative ranking probability curve for each combination of traditional Chinese medicine and Western medicine treatment regimen was calculated for each outcome indicator. The calculation formula is as follows: in, For the first A treatment plan combining traditional Chinese medicine and Western medicine was implemented in the first phase. The area under the cumulative probability ranking curve for each outcome metric. To participate in the The total number of treatment regimens evaluated by each outcome indicator. For the first A treatment plan combining traditional Chinese medicine and Western medicine was implemented in the first phase. Ranking on each outcome metric For the first A treatment plan combining traditional Chinese medicine and Western medicine was implemented in the first phase. The probability that the ranking on any outcome indicator does not exceed the mth position; The efficacy of different traditional Chinese medicine treatment regimens was ranked across various outcome indicators by the area under the cumulative probability curve, resulting in a multi-dimensional relative efficacy evaluation.

4. The clinical comprehensive evaluation method for antiviral traditional Chinese medicines according to claim 3, characterized in that, Select the corresponding effect size index according to the type of outcome index, and use the Markov chain Monte Carlo algorithm for Bayesian inference. Based on the inferred effective results, the effect size and confidence interval of each traditional Chinese medicine combined with Western medicine treatment regimen relative to Western medicine treatment alone were calculated, including: For binary outcome indicators, the odds ratio was used as the effect size. The effect size and confidence interval of each traditional Chinese medicine combined with Western medicine treatment regimen relative to Western medicine alone were calculated for each binary outcome indicator. The calculation formula is as follows: in, For the first The ratio of traditional Chinese medicine combined with Western medicine treatment to Western medicine alone. For the first The number of cases of effectiveness or adverse reactions of the combination therapy regimen For the first Total sample size of the combination therapy regimens This refers to the number of cases where Western medicine alone provides effective treatment or causes adverse reactions. This represents the total sample size for Western medicine-only treatments. for The natural logarithm of follows a normal distribution. The standard normal distribution quantiles corresponding to the 95% confidence level. for The 95% confidence interval; For continuous variable outcome indicators, the mean difference was used as the effect size. The effect size and confidence interval of each traditional Chinese medicine combined with Western medicine treatment regimen relative to Western medicine alone were calculated for each continuous variable outcome indicator. The calculation formula is as follows: in, For the first The mean difference in target continuity outcome indicators between the combined Chinese patent medicine and Western medicine treatment regimens and Western medicine treatment regimens alone. For the first The sample mean of the target continuous outcome indicators in the combination therapy regimens. This represents the sample mean of the target continuous outcome indicators in Western medicine-only treatments. For the first The sample standard deviation of the target continuity outcome measure in various combination therapy regimens. This represents the sample standard deviation of the target continuous outcome indicator in Western medicine-only treatment. for The 95% confidence interval.

5. The clinical comprehensive evaluation method for antiviral traditional Chinese medicines according to claim 1, characterized in that, The relative efficacy evaluation results and the publicly available pharmacoeconomic data are input into a pre-constructed decision tree model. The decision tree model is then used to conduct a comparative economic analysis of different traditional Chinese medicine treatment regimens to obtain economic evaluation results, including: After extracting the effect size and confidence interval of each traditional Chinese medicine (TCM) combined with Western medicine treatment regimen relative to Western medicine alone under the overall clinical efficacy outcome index, the effect size of each TCM combined with Western medicine treatment regimen relative to Western medicine treatment is converted into the absolute efficacy in the decision tree model. The calculation formula is as follows: in, For the first The absolute effectiveness of a combination of traditional Chinese medicine and Western medicine treatment in a decision tree model. The overall clinical efficacy rate was the highest among the clinical outcome indicators. The ratio of traditional Chinese medicine combined with Western medicine treatment to Western medicine alone. The baseline efficacy rate for Western medicine treatment alone. Subgroup correction factor To publish the bias correction factor, The total number of subgroups, For the first Sample size weights for each subgroup For the first The treatment in the first Efficiency in each subgroup For the first The overall effectiveness rate of this treatment To determine the number of studies included, For the first The first treatment The logarithmic effect size of each study For the first The mean of the logarithmic effect sizes from all studies on this treatment. For the first Standard error of the therapeutic effect size; The total cost of different combined traditional Chinese medicine and Western medicine treatment regimens is calculated according to cost accounting rules. The calculation formula is as follows: in, For the first The total cost of a treatment plan combining traditional Chinese medicine and Western medicine. The average daily hospitalization cost, For the first The average daily cost of using this type of traditional Chinese medicine. For the first The average daily cost of auxiliary drugs required for various traditional Chinese medicine preparations For the first The average length of hospital stay for each treatment option This represents the total number of adverse reaction types. For the first The first treatment regimen occurred The probability of adverse reactions. To process the first The average daily drug cost for each type of adverse reaction. For the reason The average daily cost of additional hospitalization due to adverse reactions. For the reason The average length of hospital stay due to adverse reactions This represents the total number of laboratory monitoring items. For the first This type of treatment requires additional monitoring. The frequency of each project For the first The cost of a single monitoring project The average length of hospital stay for Western medicine treatment alone; The absolute efficiency and the calculated total cost are input into the decision tree model for cost-effectiveness analysis, and the results are sorted from low to high cost. The calculation formula is as follows: ， ； in, The expected efficacy values ​​for each treatment regimen are given. Cost-effectiveness ratio of each treatment option Weighting of life years for clinically effective status based on quality. Weighting of quality-adjusted life years for clinically ineffective states; After eliminating the absolutely disadvantageous solutions with high costs but low outputs, calculate the incremental cost-effectiveness ratio for each pair of the remaining solutions. The calculated incremental cost-effectiveness ratio is compared with the preset willingness-to-pay threshold. If the incremental cost-effectiveness ratio is lower than the preset willingness-to-pay threshold, it indicates that the incremental cost is economical and a better treatment option should be recommended. If the incremental cost-effectiveness ratio is higher than the preset willingness-to-pay threshold, the lower-cost option will be selected first.

6. The clinical comprehensive evaluation method for antiviral traditional Chinese medicines according to claim 5, characterized in that, After eliminating the absolutely disadvantageous solutions that are high in cost but low in output, the formula for calculating the incremental cost-effectiveness ratio for pairwise comparisons of the remaining solutions is as follows: in, The incremental cost-effectiveness ratio for pairwise comparison of the remaining options. and The first The cost and effectiveness of this treatment plan and The costs and effects of the reference scheme are respectively. and These represent the variance of the cost estimate. and These represent the variance of the effect estimate.

7. The clinical comprehensive evaluation method for antiviral traditional Chinese medicines according to claim 1, characterized in that, The training set is input into a pre-built deep belief neural network model for training, a loss function is used for joint optimization, and the optimal deep belief neural network model is selected using a validation set, including: An unsupervised greedy layer-by-layer training method is adopted to pre-train each hidden layer of the deep belief network. Based on the input layer data, the feature representation of each layer is learned by a restricted Boltzmann machine (RBM). First, train the RBM consisting of the input layer and the first hidden layer. Then, use the output of the first hidden layer as the input of the second RBM for training. Continue this process to complete the pre-training of all hidden layers and initialize the model weight parameters. The pre-trained model weights are used as initial parameters and connected to the output layer to form a complete deep belief neural network. A supervised learning approach is adopted to match the input features of the training set with the corresponding real clinical outcome labels. The weights and biases of the entire network are adjusted through the backpropagation algorithm. After joint optimization using a loss function, the optimal deep belief neural network model is selected using a validation set.

8. The clinical comprehensive evaluation method for antiviral traditional Chinese medicines according to claim 1, characterized in that, The formula for calculating the loss function is: in, For loss function, The total number of samples in the training set. For sample index, For category indexing, For the first The sample at the th Real labels in each category The output of the Softmax layer The sample at the th Predict probabilities for each category. The L2 regularization coefficient is... The squared L2 norm of all weight parameters, The sparsity penalty coefficient, For hidden layer index, The total number of hidden layers in a deep confidence neural network model. KL divergence is used to constrain the average activation rate of neurons in the hidden layer. Approximate target sparsity parameter .

9. The clinical comprehensive evaluation method for antiviral traditional Chinese medicines according to claim 1, characterized in that, The formula for calculating the comprehensive value score of each traditional Chinese medicine regimen based on the relative efficacy evaluation results, economic evaluation results, and individualized efficacy prediction results is as follows: in, The overall value score for each traditional Chinese medicine prescription is calculated. The relative weights of the efficacy evaluation dimensions As the weight of the economic evaluation dimension, Weights for individualized efficacy prediction dimensions This refers to the total number of outcome indicators used in the efficacy evaluation. For the first A treatment plan combining traditional Chinese medicine and Western medicine was implemented in the first phase. The area under the cumulative probability ranking curve for each outcome metric. For the first Incremental cost-effectiveness ratio of the proposed solution to the reference solution; and These represent the maximum and minimum ICER values ​​for all included evaluation schemes. Predicting the first [number]th ... The efficacy accuracy of this treatment in the target patient population. and These represent the maximum and minimum accuracy rates of individualized efficacy prediction across all included evaluation protocols.

10. A comprehensive clinical evaluation system for traditional Chinese medicine preparations for treating respiratory viruses, characterized in that, include: The acquisition module is used to acquire literature on randomized controlled trials of traditional Chinese medicine for treating respiratory infections, real clinical case data, and publicly available pharmacoeconomic data. The relative efficacy evaluation module is used to construct an evidence network with Western medicine treatment as the common comparison basis based on the randomized controlled trial literature using the Bayesian network meta-analysis method. It calculates the effect size and confidence interval of each Chinese patent medicine combined with Western medicine treatment regimen relative to Western medicine treatment alone, and ranks the efficacy of different Chinese patent medicine treatment regimens on each outcome index by the area under the cumulative ranking probability curve, so as to obtain multi-dimensional relative efficacy evaluation results. The economic comparison analysis module is used to input the relative efficacy evaluation results and the publicly available pharmacoeconomic data into a pre-constructed decision tree model, and to conduct an economic comparison analysis of different traditional Chinese medicine treatment plans through the decision tree model to obtain economic evaluation results. The merging and partitioning module is used to merge the real clinical case data, relative efficacy evaluation results, and economic evaluation results data and then divide them into training set, test set, and validation set according to a certain ratio. The training module is used to input the training set into a pre-built deep belief neural network model for training, perform joint optimization using a loss function, and select the optimal deep belief neural network model using a validation set. The efficacy prediction module is used to input the test set into the optimal deep belief neural network model and output individualized efficacy prediction results. The comprehensive evaluation module is used to calculate the comprehensive value score of each traditional Chinese medicine prescription based on the relative efficacy evaluation results, economic evaluation results, and individualized efficacy prediction results, and to evaluate the traditional Chinese medicine prescriptions against respiratory viruses through the comprehensive evaluation score.

Images