BACKGROUND:Due to the complex movement of the scapula,which is a six-degree-of-freedom activity in three-dimensional space,it is difficult to measure it accurately using traditional methods.The image and model matching technology based on dual-plane X-ray is a three-dimensional measurement method that has gradually developed and matured in recent years.Two high-speed cameras are used to project and shoot from orthogonal directions.Compared with a single perspective,this method has advantages in observation range and reduction of out-of-plane errors,and is suitable for the study of scapula kinematics.OBJECTIVE:X-ray biplane and image-model registration technology were used to explore the differences in scapular kinematics between normal individuals and patients with rotator cuff tears,providing a basis for the treatment and rehabilitation of rotator cuff tear patients.METHODS:From April 2023 to January 2024,10 patients with normal shoulders and 10 patients with rotator cuff tears who met the inclusion criteria were enrolled from Guanghua Hospital Affiliated to Shanghai University of Traditional Chinese Medicine.The subjects underwent a shoulder CT scan to create a 3D model and a local scapular coordinate system.X-ray biplane images were taken during shoulder abduction with two C-arm machines.The 2D and 3D images were registered to compare scapular kinematic differences between the two groups at 0°,15°,30°,45°,60°,75°,and 90° of abduction,including scapular rotation angle and displacement distance.RESULTS AND CONCLUSION:(1)During shoulder abduction,the scapula of both groups showed upward rotation,but the upward rotation of the rotator cuff tear group was greater than that of the control group,and the difference was significant when the abduction was 30°-90°(P＜0.01).At the same time,the scapula internal rotation of both groups gradually increased,but the rotator cuff tear group was significantly greater than the control group when the abduction was 45°-90°(P＜0.01).In addition,the anteroposterior tilt of the scapula of the two groups was significantly different when the abduction was 15°-90°(P＜0.01).The scapula posterior tilt of the control group gradually increased during abduction,while the scapula of the rotator cuff tear group tilted forward except for a slight posterior tilt at 15°-30° abduction.(2)In terms of displacement,the upward displacement of the rotator cuff tear group was less than that of the control group during abduction,and the difference was significant at 15°-90°(P＜0.05),but there was no significant difference in lateral and anterior-posterior displacement between the two groups(P＞0.05).(3)Rotator cuff tear can cause scapular dyskinesis,characterized by increased upward rotation,internal rotation,and abnormal forward tilt during shoulder abduction.Identifying and addressing scapular dyskinesis is crucial for treating rotator cuff tear.

OBJECTIVE:Based on different algorithms of machine learning,the prediction model of lumbar disc herniation has become a trend and hot spot in the development of precision medicine.However,there is limited evidence on the reporting quality and methodological quality of prediction models of lumbar disc herniation outcomes using machine learning.This article is aimed to explore the performance of machine learning algorithms in predicting the prognosis of lumbar disc herniation by comprehensively analyzing the report quality and risk of bias of previous studies that developed and validated prognosis prediction models based on machine learning through a comprehensive literature search,in order to explore the performance of machine learning algorithms in predicting the prognosis of lumbar disc herniation.METHODS:The databases of CNKI,WanFang,VIP,SinOMED,PubMed,Web of Science,Embase,and The Cochrane Library were searched by computer.Studies on the use of machine learning to develop(and/or validate)prognostic prediction models for lumbar disc herniation were collected from the inception of the database to December 31,2023.Two researchers independently screened the literature,extracted data,and assessed the risk of bias of the included studies.The reporting quality and risk of bias of the included studies were assessed by the Multivariable Transparent Reporting of Predictive Models(TRIPOD)statement and the Predictive Model Risk of Bias Assessment Tool(PROBAST).The results of the evaluation were analyzed using descriptive statistics and visual charts.RESULTS:(1)A total of 23 articles were included,and the TRIPOD compliance of each study ranged from 11％to 87％,with a median compliance of 54％.The quality of reporting of titles,detailed descriptions of treatment measures,blinding of predictors,handling of missing data,details of risk stratification,specific procedures for enrollment,model interpretation,and model performance was mostly poor,with TRIPOD adherence rates ranging from 4％to 35％.(2)Of all included studies,61％had a high risk of bias and 39％had an unclear overall risk of bias.The area under the curve,accuracy,sensitivity and specificity were used to evaluate the performance of the model.The areas under the curve of 20 models were reported,ranging from 0.561 to 0.999.Three models reported the accuracy of the model,ranging from 82.07％to 89.65％.(3)Among all included studies,the statistical analysis domain was most often assessed as having a high risk of bias,mainly due to the small number of valid samples,the selection of predictors based on univariate analysis and the lack of calibration and discrimination assessment of the model in the study.CONCLUSION:These results indicate that machine learning can achieve good predictive ability in the development and validation of prognostic models for lumbar disc herniation.The commonly used algorithms include regression algorithm,support vector machine,decision tree,random forest,artificial neural network,naive Bayes and other algorithms.Reasonable algorithms combined with clinical practice can improve the accuracy of prognosis prediction of lumbar disc herniation.However,the reporting and methodological quality of prognosis prediction models based on machine learning are poor,the prediction performance of different models varies greatly,and the generalization and extrapolation of research models are unclear.There is an urgent need to improve the design,implementation and reporting of such studies.To promote the application of machine learning in the clinical practice of lumbar disc herniation prediction models,it is necessary to comprehensively consider various predictors related to the prognosis of the disease before modeling,and strictly follow the relevant standards of PROBAST tool during modeling.

Objective To investigate the influencing factors of stroke patients with hypertension and their compliance with antihypertensive drugs, and to provide targeted intervention measures for stroke prevention in hypertensive patients.  Methods Using the method of multi-stage cluster sampling, a total of 59,434 permanent residents aged 40 and above were selected from 48 monitoring sites in 9 cities of Hebei Province from December 2019 to December 2020. Unconditional logistic regression analysis was used to explore the influencing factors of stroke complicated with hypertension and the compliance of patients with antihypertensive drugs.  Results Among the 59 434 subjects, the prevalence rate of stroke was 4.33% and the prevalence rate of stroke complicated with hypertension was 82.47%. The results of univariate analysis showed that the proportion of women, rural areas, dyslipidemia, diabetes, obesity, and people with family history of stroke was higher in stroke patients with hypertension, and the difference was statistically significant (P<0.05). Multivariate logistic regression analysis showed that rural areas, dyslipidemia, diabetes, obesity, and family history of stroke significantly increased the risk of stroke, and the OR (95%CI) values were 1.29 (1.03-1.62), 1.39 (1.12-1.72), 1.58 (1.25-1.99), 1.61 (1.22-2.12) and 1.60 (1.26-2.04), respectively. Among stroke patients with hypertension, 92.71% of patients took antihypertensive drugs. It was found that women's compliance with antihypertensive drugs was good, with an OR (95%CI) value of 1.46 (1.01-2.09).  Conclusion The prevalence rate of stroke complicated with hypertension is high in people aged 40 and above in Hebei Province. Hypertensive people should lower blood lipids, control blood glucose, and lose weight as soon as possible to prevent the occurrence of stroke.

OBJECTIVE To evaluate the efficacy， safety， and cost-effectiveness of posaconazole for the prevention of invasive fungal infections （IFIs）. METHODS PubMed， Embase， CNKI， Wanfang data and other Chinese and English databases were searched， as well as the official websites of related health technology assessment （HTA）. Systematic review （SR）/meta-analysis， pharmacoeconomic studies， and HTA reports of posaconazole for the prevention of IFIs were collected. The retrieval time limit was from the establishment of the database to November 1， 2025. After screening the literature， extracting the data， and evaluating the quality of the literature， the results of the included studies were descriptively analyzed. RESULTS A total of 45 studies were included， involving 17 SR/meta-analyses， 26 pharmacoeconomic studies， 1 SR/meta-analysis combined with a pharmacoeconomic study， and 1 HTA report. In terms of effectiveness， posaconazole was significantly superior to other antifungal drugs （eg.， fluconazole， itraconazole） in reducing the incidence of IFIs， the incidence of invasive aspergillosis， all-cause mortality， and IFI-related mortality （ P ＜0.05）. In terms of safety， posaconazole was significantly superior to other antifungal drugs in the incidence of total adverse reactions （ P ＜0.05）， but there was no significant difference in the incidence of serious adverse reactions， the incidence of gastrointestinal advers e reactions， drug-induced liver injury， and the withdrawal rate due to adverse drug reactions （ P ＞0.05）. In terms of cost-effectiveness， most studies have shown that posaconazole possesses more cost-effectiveness advantages. CONCLUSIONS Posaconazole demonstrates favorable efficacy， safety， and cost-effectiveness in preventing IFIs.

ObjectiveTo explore the efficacy and mechanism of Euphorbia humifusa on acute kidney injury （AKI） based on network pharmacology， molecular docking and experimental verification. MethodsThe active components and targets of E. humifusa were retrieved from TCMSP and SwissTargetPrediction database， and the AKI targets were screened by GeneCards and Online Mendelian Inheritance in Man（OMIM） databases. The drug targets and disease targets were intersected to construct a protein-protein interaction network， and the intersection targets were subjected to gene ontology （GO） and Kyoto encyclopedia of genes and genomes （KEGG） enrichment analysis. Discover Studio software was used to verify the molecular docking of key components and core targets. Gentamicin （GM） was used to induce AKI rat model. Control group， model group， verapamil （16 mg·kg^-1） group， E. humifusa extract （18， 54， 162 mg·kg^-1·d^-1） group and E. humifusa 70% ethanol extract （423 mg·kg^-1） group were continuously administered for 14 days. Urine volume was detected 24 h after modeling and administration. Serum creatinine （SCr）， Blood urea nitrogen （BUN）， 24-hour urine protein （24 hUTP） and uric acid （UA） content； the contents of malondialdehyde （MDA）， glutathione （GSH）， superoxide dismutase （SOD）， carbon monoxide synthase （NOS） and lactate dehydrogenase （LDH） in kidney were measured. The levels of interleukin （IL）-6 and tumor necrosis factor （TNF）-α in serum were detected by enzyme linked immunosorbent assay（ELISA） kit. The pathological changes of renal tissue were detected by hematoxylin-eosin （HE） and Masson staining. Western blot was used to detect the expression of PI3K/protein kinase B（Akt）/NF-κB signaling pathway-related proteins. ResultsIn this study， 13 active components such as kaempferol， luteolin， apigenin， gallic acid and quercetin were screened and identified from E. humifusa. Through bioinformatics analysis， these components and AKI have a total of 289 targets， of which 62 are core targets， including Akt1， TNF， tumor protein p53（TP53） and IL-1β. These targets are mainly involved in the regulation of biological processes such as NF-κB signaling pathway， HIF-1 signaling pathway， TNF signaling pathway， PI3K/Akt signaling pathway and mitogen-activated protein kinase（MAPK） signaling pathway. In animal experiments， we successfully constructed a GM-induced AKI model in rats. Compared with the model group， E. humifusa extract could significantly reduce the levels of 24 hUTP， BUN and SCr in rats （P<0.01）， indicating its improvement effect on renal function. In addition， the extract of E. humifusa also significantly reduced LDH activity and MDA content in rat kidney tissue （P<0.05， P<0.01）， and significantly increased SOD， NOS activity and GSH content （P<0.05）， indicating that the extract of E. humifusa has the potential of anti-oxidation and protection of renal function. Further analysis of inflammatory factors showed that the levels of IL-6 and TNF-α in serum of rats treated with E. humifusa extract were significantly decreased （P<0.01）， indicating that E. humifusa extract had anti-inflammatory effects. In addition， the extract of E. humifusa can also regulate the protein expression of PI3K/Akt/NF-κB signaling pathway， which further confirmed its mechanism of reducing GM-induced AKI. ConclusionThe extract of E. humifusa has a significant therapeutic effect on acute kidney injury through its multi-component and multi-target mechanism. Its effect is reflected in improving renal function， anti-oxidation， anti-inflammation and regulating immune response. These findings provide a scientific basis for the application of E. humifusa in the treatment of acute kidney injury， and point out the direction for future drug development and clinical research.

ObjectiveThis paper aims to investigate the efficacy and related actions of Guizhi Fulingwan in intervening in the mice with colitis-associated colon cancer （CAC） based on the immunosuppressive microenvironment associated with myeloid-derived suppressor cells （MDSCs）. MethodsSixty male C57BL/6 mice were randomly assigned to a blank group， a model group， an aspirin group （0.04 g·kg^-1）， and low-， medium-， and high-dose Guizhi Fulingwan groups （4.87， 9.75， and 19.50 g·kg^-1）， with ten mice per group. The CAC mouse model was established via combined induction of azoxymethane （AOM）/dextran sulphate sodium （DSS）. Drug intervention commenced in week five， with continuous intragastric administration for nine weeks. The food intake， body weight， fecal characteristics， and haematochezia were observed and recorded， and disease activity index （DAI） scores were calculated according to scoring criteria. Hematoxylin and eosin （HE） staining was used to observe the histopathological changes in the colon tissues of the mice. Immunohistochemistry was used to determine proliferating cell nuclear antigen-67 （Ki67） expression in the colon tissues， and enzyme-linked immunosorbent assay （ELISA） was used to detect the contents of interleukin-6 （IL-6）， IL-1β， and tumor necrosis factor-α （TNF-α） in the serum of the mice. Flow cytometry was employed to determine the proportion levels of MDSCs， CD4⁺ T cells， and CD8⁺ T cells in the spleen tissues of the mice. The mRNA expressions of MDSC-associated effector molecules， including arginase 1 （Arg1） and inducible nitric oxide synthase （iNOS）， were detected by real-time quantitative polymerase chain reaction （Real-time PCR）. After that， an in vitro co-culture model of MDSCs and CD8⁺ T cells was established， and drug-containing serum of Guizhi Fulingwan was used for intervention. The Flow cytometry was employed to assess the effects of drug-containing serum of Guizhi Fulingwan with different concentrations on the levels of reactive oxygen species （ROS） and iNOS in MDSCs and the proliferation of CD8⁺ T cells. The levels of granzyme B （GZMB） and interferon-γ （IFN-γ） in cell supernatant were detected by ELISA. ResultsCompared with those in the control group， the mice in the model group exhibited significantly reduced body weight， elevated DAI scores， shortened colon length （P<0.01）， increased number of tumors and Ki67 expression （P<0.01）， and significantly elevated contents of IL-6， IL-1β， and TNF-α in the serum （P<0.01）. Significant increases in the number of MDSCs were observed in mouse spleens， alongside marked reductions in the levels of CD4⁺ T and CD8⁺ T cells （P<0.01）. Furthermore， the mRNA expressions of MDSC function-associated effector molecules Arg1 and iNOS were significantly upregulated （P<0.01）. Compared with those in the model group， the mice in the middle-dose Guizhi Fulingwan group exhibited increased body weight and significantly decreased DAI scores （P<0.05， P<0.01）. The mice in the middle- and high-dose Guizhi Fulingwan groups exhibited significantly improved colon shortening， significantly decreased number of tumors and Ki67 expression （P<0.05， P<0.01）， and significantly decreased contents of IL-6， IL-1β， and TNF-α in the serum （P<0.05， P<0.01）. Furthermore， administration of Guizhi Fulingwan markedly reduced MDSC infiltration in the spleen of the mice， with different degrees of increase in the levels of both CD4⁺ T and CD8⁺ T cells （P<0.05， P<0.01）， alongside significant decreases in the mRNA expressions of Arg1 and iNOS （P<0.05， P<0.01）. In vitro cell co-culture shows that administration of drug-containing serum of Guizhi Fulingwan significantly decreases the activity levels of ROS and iNOS in MDSCs and promotes the proliferation of CD8⁺ T cells and the secretion of GZMB and IFN-γ （P<0.05， P<0.01）. ConclusionGuizhi Fulingwan can reduce pro-inflammatory cytokine secretion and inhibit tumor proliferation in the colon tissues of CAC mice. Its potential mechanism may involve reducing MDSC infiltration， enhancing effector T cells， particularly CD8⁺ T cell response， and improving the tumor immunosuppressive microenvironment.

Objective: To investigate the correlation between the color parameters (L^*, a^*, b^*, Eab^* values) of Polygoni Multiflori Radix Praeparata powder prepared by different processing techniques and the contents of 2,3,5,4’-tetrahydroxystilbene-2-O-β-D-glucopyranoside, emodin, physcion, emodin-8-O-β-D-glucopyranoside, physcion-8-O-β-D-glucopyranoside.
Methods: The L^*, a^*, b^* and Eab^* values of Polygoni Multiflori Radix Praeparata powder prepared by different processing techniques were determined by spectrophotometer, and the contents of the five components were determined by high performance liquid chromatography. Secondly, SPSS 26.0 software and Simca 14.1 software were used to analyze the correlation.
Results: Through the hierarchical cluster analysis (HCA), it was found that the steamed samples and black bean steamed samples could be obviously divided into two categories: raw products and processed products. The processed products could be further divided into 2-8 h and 12-48 h. Pearson correlation analysis showed that the content of stilbene glycoside was significantly positively correlated with L^*, a^* and b^* values (P<0.01). The a^* value was significantly positively correlated with the content of emodin and physcion (P<0.01). Emodin-8-O-β-D-glucoside was positively correlated with L^* value and b^* value, while physcion-8-O-β-D-glucoside was negatively correlated with a^* value. Partial least squares discriminant analysis (PLS) showed that 2,3,5,4’-tetrahydroxystilbene-2-O-β-D-glucoside (VIP=1.69) and emodin-8-O-β-D-glucoside (VIP=1.06) were the key variables affecting chromaticity characteristics (P<0.01).
Conclusion: The three processes of steaming, black bean steaming and black bean stewing are consistent in composition transformation and chromaticity variation, and stilbene glycoside can be used as a specific index component to characterize the processed color. Chromatic parameters can effectively reflect the processing progression and serve as quality monitoring indicators during production.

BACKGROUND:Due to the complex movement of the scapula,which is a six-degree-of-freedom activity in three-dimensional space,it is difficult to measure it accurately using traditional methods.The image and model matching technology based on dual-plane X-ray is a three-dimensional measurement method that has gradually developed and matured in recent years.Two high-speed cameras are used to project and shoot from orthogonal directions.Compared with a single perspective,this method has advantages in observation range and reduction of out-of-plane errors,and is suitable for the study of scapula kinematics.OBJECTIVE:X-ray biplane and image-model registration technology were used to explore the differences in scapular kinematics between normal individuals and patients with rotator cuff tears,providing a basis for the treatment and rehabilitation of rotator cuff tear patients.METHODS:From April 2023 to January 2024,10 patients with normal shoulders and 10 patients with rotator cuff tears who met the inclusion criteria were enrolled from Guanghua Hospital Affiliated to Shanghai University of Traditional Chinese Medicine.The subjects underwent a shoulder CT scan to create a 3D model and a local scapular coordinate system.X-ray biplane images were taken during shoulder abduction with two C-arm machines.The 2D and 3D images were registered to compare scapular kinematic differences between the two groups at 0°,15°,30°,45°,60°,75°,and 90° of abduction,including scapular rotation angle and displacement distance.RESULTS AND CONCLUSION:(1)During shoulder abduction,the scapula of both groups showed upward rotation,but the upward rotation of the rotator cuff tear group was greater than that of the control group,and the difference was significant when the abduction was 30°-90°(P＜0.01).At the same time,the scapula internal rotation of both groups gradually increased,but the rotator cuff tear group was significantly greater than the control group when the abduction was 45°-90°(P＜0.01).In addition,the anteroposterior tilt of the scapula of the two groups was significantly different when the abduction was 15°-90°(P＜0.01).The scapula posterior tilt of the control group gradually increased during abduction,while the scapula of the rotator cuff tear group tilted forward except for a slight posterior tilt at 15°-30° abduction.(2)In terms of displacement,the upward displacement of the rotator cuff tear group was less than that of the control group during abduction,and the difference was significant at 15°-90°(P＜0.05),but there was no significant difference in lateral and anterior-posterior displacement between the two groups(P＞0.05).(3)Rotator cuff tear can cause scapular dyskinesis,characterized by increased upward rotation,internal rotation,and abnormal forward tilt during shoulder abduction.Identifying and addressing scapular dyskinesis is crucial for treating rotator cuff tear.

OBJECTIVE:Based on different algorithms of machine learning,the prediction model of lumbar disc herniation has become a trend and hot spot in the development of precision medicine.However,there is limited evidence on the reporting quality and methodological quality of prediction models of lumbar disc herniation outcomes using machine learning.This article is aimed to explore the performance of machine learning algorithms in predicting the prognosis of lumbar disc herniation by comprehensively analyzing the report quality and risk of bias of previous studies that developed and validated prognosis prediction models based on machine learning through a comprehensive literature search,in order to explore the performance of machine learning algorithms in predicting the prognosis of lumbar disc herniation.METHODS:The databases of CNKI,WanFang,VIP,SinOMED,PubMed,Web of Science,Embase,and The Cochrane Library were searched by computer.Studies on the use of machine learning to develop(and/or validate)prognostic prediction models for lumbar disc herniation were collected from the inception of the database to December 31,2023.Two researchers independently screened the literature,extracted data,and assessed the risk of bias of the included studies.The reporting quality and risk of bias of the included studies were assessed by the Multivariable Transparent Reporting of Predictive Models(TRIPOD)statement and the Predictive Model Risk of Bias Assessment Tool(PROBAST).The results of the evaluation were analyzed using descriptive statistics and visual charts.RESULTS:(1)A total of 23 articles were included,and the TRIPOD compliance of each study ranged from 11％to 87％,with a median compliance of 54％.The quality of reporting of titles,detailed descriptions of treatment measures,blinding of predictors,handling of missing data,details of risk stratification,specific procedures for enrollment,model interpretation,and model performance was mostly poor,with TRIPOD adherence rates ranging from 4％to 35％.(2)Of all included studies,61％had a high risk of bias and 39％had an unclear overall risk of bias.The area under the curve,accuracy,sensitivity and specificity were used to evaluate the performance of the model.The areas under the curve of 20 models were reported,ranging from 0.561 to 0.999.Three models reported the accuracy of the model,ranging from 82.07％to 89.65％.(3)Among all included studies,the statistical analysis domain was most often assessed as having a high risk of bias,mainly due to the small number of valid samples,the selection of predictors based on univariate analysis and the lack of calibration and discrimination assessment of the model in the study.CONCLUSION:These results indicate that machine learning can achieve good predictive ability in the development and validation of prognostic models for lumbar disc herniation.The commonly used algorithms include regression algorithm,support vector machine,decision tree,random forest,artificial neural network,naive Bayes and other algorithms.Reasonable algorithms combined with clinical practice can improve the accuracy of prognosis prediction of lumbar disc herniation.However,the reporting and methodological quality of prognosis prediction models based on machine learning are poor,the prediction performance of different models varies greatly,and the generalization and extrapolation of research models are unclear.There is an urgent need to improve the design,implementation and reporting of such studies.To promote the application of machine learning in the clinical practice of lumbar disc herniation prediction models,it is necessary to comprehensively consider various predictors related to the prognosis of the disease before modeling,and strictly follow the relevant standards of PROBAST tool during modeling.

ObjectiveTo systematically evaluate the content differences of 4 components in Leonuri Herba granules， reveal the quality fluctuation patterns of products from the same and different manufacturers， providing scientific basis for the optimization of production process and quality control. MethodsHigh performance liquid chromatography-diode array detector-charged aerosol detector（HPLC-DAD-CAD） was employed to determine the contents of 4 components（syringic acid， leonurine hydrochloride， ferulic acid， and stachydrine hydrochloride） in samples from 19 manufacturers（53 batches， 159 boxes）. Additionally， fingerprint profiles were constructed， and the fingerprint dissimilarity（P_S） and relative standard deviation（RSD） of different samples from the same manufacturer were calculated. A principal component analysis（PCA） model was established with P_S and the RSD values of the 4 components as variables to classify the manufacturers. Finally， samples from 5 manufacturers（M1-M5） covering three consistency groups were selected to calculate three quality consistency parameters， namely intra-batch consistency（P_A）， inter-batch consistency（P_B）， and P_S. Then， PCA was performed with P_A， P_B， and P_S of these 5 manufacturers as variables. ResultsThe average total content of the 4 index components per bag across the 19 manufacturers ranged from 41.10 mg to 97.54 mg. Among them， the content of stachydrine hydrochloride（a pharmacopoeial quality control component） was 32.46-72.70 mg per bag， all meeting the requirements of the 2025 edition of the Pharmacopoeia of the People's Republic of China， with RSD of 1.7%-17.1%. The content ranges of the other 3 components were as follows：syringic acid of 1.43-41.92 mg per bag， leonurine hydrochloride of 0.67-11.85 mg per bag， and ferulic acid of 0.11-3.81 mg per bag. Notably， leonurine hydrochloride exhibited the most significant content fluctuation among samples from the same manufacturer（RSD of 4.8%-59.2%）. PCA results showed that the 19 manufacturers could be classified into 3 categories. Samples from 8 manufacturers（M2， M6， M7， M8， M10， M15， M17， M18） demonstrated relatively high consistency， five manufacturers（M3， M9， M12， M13， M14） showed moderate consistency， six manufacturers（M1， M4， M5， M11， M16， M19） exhibited low consistency. The two methods yielded consistent classification results for the 5 representative manufacturers， verifying the reliability of the proposed method. Among these， manufacturer M2 showed the best quality consistency and the highest total content of indicator components among M1-M5. ConclusionThe HPLC-DAD-CAD multi-detector hyphenation technology established in this study enables the accurate detection of 4 components in Leonuri Herba granules. Significant differences in the total content of these four components are observed among products from 19 manufacturers. The application of 2 consistency evaluation methods combined with PCA can effectively classify their consistency into 3 categories， and the classification results of the 2 methods are highly consistent. This study provides scientific basis for the process optimization and quality standard improvement of Leonuri Herba granules.