[Objective]To examine the botanical origin of the exotic medicinal materia Alebo firstly contained in Bencao Shiyi.[Methods]Based on the records of Haiyao Bencao,Compendium of Materia Medica,etc.,through ancient and modern literature verification and comparative botanical studies,combined with fieldwork in the field,from the perspectives of plant morphology,habitat distribution,herb name,nature and flavor,efficacy,the Alebo was examined.[Results]Comparative study revealed that the Alebo(Brahman saponaria)of the Bencao Shiyi matches the morphological characteristics of the Leguminosae Tamarindus indica L.(T.indica).T.indica are native to the Indian Peninsula and the Malay Archipelago,matching the habitat and name of the Brahmin saponaria.The two properties of similar nature and flavor,efficacy and mouth taste of T.indica peel and seeds are in line with the Brahmin Saponaria of the"taste".The Compendium of Materia Medica blackmail variant of Abole(Persian saponaria)is morphologically consistent with the Leguminosae Cassia fistula L.(C.fistula),and also bears some of the characteristics of the Kigelia africana(Lam.)Benth.(K.africana).In conjunction with further research,the habitat,name,and nature and flavor,efficacy effects of the C.fistula are consistent with the characteristics of Abole.Compendium of Materia Medica appendix Guihai Zhi tamarind,Sterculiaceae Sterculia nobilis Smith(S.nobilis)in line with the"nine layers of skin fruit"one.S.nobilis and Sterculia lanceolata Cav.(S.lanceolata)are consistent with tamarind in morphology,habitat,and nature and flavor,efficacy.[Conclusion]It is suggested that in the revision of the Dictionary of Chinese Medicine and other books,the origin of"Brahmin saponaria"should be changed to T.indica of the Leguminosae,and that in the revision of the Chinese Materia Medica Uyghur Medicine Volume and other books,the origin of"tamarind"should be changed to S.nobilis and S.lanceolata of the Sterculiaceae.In order to correct the origin,renew the history of medicine,inherit the essence.

Objective:To explore the efficacy of the features of ultrasound imaging omics at the area peripheral thyroid nodule in predicting cervical lymph node metastasis(LNM)of papillary thyroid carcinoma(PTC),and construct a prediction model based on the features of imaging omics and to verify its performance.Methods:A total of 237 PTC patients who admitted to Shaanxi Provincial Hospital of Chinese Medicine from March 2021 to June 2024 and were confirmed by pathology were retrospectively collected.They were divided into the training set(166 cases)and the validation set(71 cases)as a ratio of 7 to 3.According to the postoperatively pathological results,237 patients were divided into the metastasis group(108 cases)and the non-metastasis group(129 cases).The clinical data and conventional ultrasound characteristic information of all patients were collected,and a feature model of imaging omics was constructed through quantitative extracting and screening the features of ultrasound imaging omics within nodules and peripheral nodules,and utilizing machine learning classifier.Then,the feature score(Rad-Score)of image omics was obtained.The Rad-Score values within and peripheral nodules,and the Rad-Score values peripheral nodules of metastasis group and non-metastasis group were compared.In training set,the independent risk factors of affecting neck LNM were analyzed,and a clinical-ultrasonic model was constructed,which was combined with Rad-Score to construct a joint model based on the features of imaging omics peripheral nodules.The receiver operating characteristic(ROC)curve was used to analyze and compare its predictive efficacy.The nomogram of the joint model was constructed,and then,the calibration and fitting degrees of the nomogram were assessed by calibration curve and Hosmer-Lemeshow test.Result:In training set,6 features of imaging omics within nodules and 11 features of imaging omics peripheral nodules were respectively extracted and screened out through the least absolute shrinkage and selection operator(LASSO)algorithm.In the training set and validation set,the Rad-Scores peripheral the nodules in the metastasis group were respectively(7.43±0.45)points and(7.19±0.51)points,which were significantly higher than(3.25±0.28)points and(3.51±0.32)points peripheral the nodules in the non-metastasis group,and the differences were statistically significant(t=72.708,61.222,P<0.05).The results of factor analysis showed that age,capsule invasion,microcalcification,ultrasound-indicated lymph node positivity and Rad-Score around nodules were independent risk factors of affecting cervical LNM of PTC patients(OR=0.592,2.983,3.593,4.424,2.575,P<0.05).The ROC curve showed that the area under curve(AUC)values of the ROC curve of joint model in training set and validation set were respectively 0.861 and 0.872 in predicting LNM,respectively,which were superior to 0.759 and 0.783 of the clinical-ultrasound model.Conclusion:In both the training set and the validation set,the nomogram of joint model has favorable calibration and fitting in predicting cervical LNM of PTC patients.The construction of clinical model based on the features of ultrasound imaging omics peripheral nodules has a favorable efficacy in predicting the status of cervical lymph node of PTC patients before surgery,which is expected to be an effective tool of individual prediction for LNM.

Objective To systematically evaluate the published models in prediction of the risk of lung infections in elderly patients with hip fracture so as to provide a guidance for medical workers in selection or development of suitable risk prediction models.Methods Relevant studies were searched from databases including CNKI,Wanfang Data,VIP,SinoMed,PubMed,Web of Science,Cochrane Library,Embase and CINAHL,from the inception to 31st January,2024.Data were extracted from the selected literature and a bias assessment tool of risk predictive model was used to evaluate the risk of bias and applicability of the included literature.Results A total of 1,035 articles were retrieved,of which seven studies involving 13 predictive models were finally included after screening.The sample sizes ranged from 305 to 2,669 cases and lung infection rates ranged from 5.40%to 20.02%.The repeatedly reported predictors included age,gender,chronic obstructive pulmonary disease,hypoproteinaemia,American Society of Anesthesiologists(ASA)Physical Status Classification and white blood cell count.In the 13 models constructed,the reported area under the curve(AUC)of subjects'job characteristics ranged from 0.667 to 0.996.Five out of seven studies had good overall applicability,but all with high risk of bias.Conclusion The predictive models for lung infections in elderly patients with hip fracture are still in the stage of development.Although the predictive models show some predictive performance,however they are still deficient,and all studies have been found with a high risk in bias.

Objective:To investigate the application value of risk prediction model for acute kidney injury (AKI) after donation of cardiac death (DCD) liver transplantation based on machine learning algorithm.Methods:The retrospective cohort study was conducted. The clinicopathological data of 1 001 pairs of DCD liver transplant donors and recipients at five hospitals, including The First Affiliated Hospital of Zhejiang University School of Medicine et al, in the Chinese Liver Transplan-tation Registry from January 2015 to December 2023 were collected. Of the donors, there were 825 males and 176 females. Of the recipients, there were 806 males and 195 females, aged 52 (range, 18-75)years. There were 281 recipients included using oversampling technique, and all 1 282 recipients were divided to the training set of 897 recipients and the validation set of 385 recipients by a ratio of 7∶3 using computer-generated random numbers. Seven prediction models, including Random Forest (RF), Extreme Gradient Boosting (XGBoost), Support Vector Machine (SVM), Logistic Regression (LR), Decision Tree (DT), K-Nearest Neighbors (KNN), and Categorical Boosting (CatBoost), were constructed for AKI after liver transplantation based on machine learning algorithm. Observation indicators: (1) comparison of clinicopathological characteristics between recipients with and without AKI and donors; (2) follow-up and survival of recipients with and without AKI; (3) construction and validation of nomogram prediction model of AKI after liver transplantation; (4) construction and validation of machine learning prediction model of AKI after liver transplantation. Comparison of measurement data with normal distribution between groups was conducted using the independent sample t test. Comparison of measurement data with skewed distribution between groups was conducted using the Mann-Whitney U test, and comparison among groups was conducted using the Kruskal-Wallis H test. Comparison of count data between groups was conducted using the chi-square test or corrected chi-square test. Kaplan-Meier method was used to calculate survival rates and plot survival curves. Logistic regression model was performed for univariate and multivariate analyses. The receiver operating characteristic (ROC) curve was plotted to calculate area under curve (AUC) and 95% confidence interval ( CI). The performance of prediction model was evaluated using DeLong test, accuracy, sensitivity, specificity. The calibration curve was plotted to evaluate the performance of predicted probability and actual probability. The interpretability analysis of machine learning algorithm and SHapley Additive exPlanations was used to explain the model decision separately. Results:(1) Comparison of clinicopathological characteristics between recipients with and without AKI and donors. Of 1 001 recipients, there were 360 cases with AKI and 641 cases without AKI after liver transplantation. There were significant differences in body mass index (BMI), hepatic encepha-lopathy, hepatitis B surfact antigen (HBsAg), hepatorenal syndrome (HRS) and donor diabetes, donor blood urea nitrogen, donor alanine aminotransferase, donor aspartate aminotransferase, mass of graft, volume of blood loss during liver transplantation, warm ischema time of donor liver, and operation time between recipients with and without AKI ( Z=-4.337, χ2=9.751, 9.088, H=11.142, χ2=5.286, Z=-3.360, -2.539, -3.084, -1.730, -3.497, -1.996, -2.644, P<0.05). (2) Follow-up and survival of recipients with and without AKI. All the 1 001 recipients received follow-up. The recipients with AKI after liver transplantation were followed up for 18.6(range, 0-102.3)months, and recipients without AKI after liver transplantation were followed up for 31.9(range, 0.1-105.5)months. The 1-, 3-, and 5-year overall survival rates were 72.1%, 63.5%, and 59.3% of recipients with AKI, versus 86.7%, 76.7%, and 72.5% of recipients without AKI, respectively, showing a significant difference in overall survival between them ( χ2=26.028, P<0.05). (3) Construction and validation of nomogram predic-tion model of AKI after liver transplantation. Results of multivariate analysis showed that recipient BMI, recipient creatinine, recipient HBsAg, recipient HRS, donor blood urea nitrogen, donor crea-tinine, anhepatic phase and volume of blood loss during liver transplantation were independent risk factors for AKI of recipients after liver transplantation ( odds ratio=1.113, 0.998, 0.605, 1.580, 1.047, 0.998, 1.006, 1.157, 95% CI as 1.070-1.157, 0.996-1.000, 0.450-0.812, 1.021-2.070, 1.021-1.074, 0.996-0.999, 1.000-1.012, 1.045-1.281, P<0.05). The nomogram prediction model of AKI after liver transplantation was constructed based on the results of multivariate analysis. Results of ROC curve showed that the AUC of 0.666 (95% CI as 0.637-0.696). (4) Construction and validation of machine learning prediction model of AKI after liver transplantation. Based on the Lasso regression analysis, seven machine learning algorithm prediction models, including RF, XGBoost, SVM, LR, DT, KNN, and CatBoost, were constructed, with ROC curves of the validation set plotted. The AUC of above models were 0.863, 0.841, 0.721, 0.637, 0.620, 0.708, 0.731, accuracies were 0.764, 0.782, 0.701, 0.592, 0.605, 0.605, 0.681, sensitivities were 0.764, 0.789, 0.719, 0.588, 0.694, 0.694, 0.704, specificities were 0.763, 0.774, 0.683, 0.597, 0.511, 0.511, 0.656, respectively. Delong test showed that the RF model with the highest AUC of 0.863(95% CI as 0.828-0.899). Calibration curve analysis showed the predicted probability closest to the actual probability of RF model, indicating the model with a good validation value. Further sorting of SHAP of different clinical factors based on RF model showed that recipient BMI, donor blood urea nitrogen, volume of blood loss during liver transplantation, donor age had large effects on the output outcomes. Conclusion:The nomogram prediction model and seven machine learning algorithm prediction models for AKI after DCD liver transplantation are constructed, and the RF model based on machine learning has a better predictive performance.

Objective To identify determinants of subtherapeutic voriconazole(VRCZ)concentrations in allogeneic hematopoietic stem cell transplantation(allo-HSCT)recipients and to develop/validate a nomogram-based risk prediction model.Methods This study retrospectively analyzed 310 VRCZ therapeutic drug monitoring(TDM)measurements from allo-HSCT recipients at 310 patients who under went allo-HSCT surgery at Hebei Yanda Ludaopei Hospital from October 2022 to October 2024 and received VRCZ for the prevention and treatment of invasive fungal infections before transplantion were selected as the study subjects.Cases were stratified into target-concentration group(0.5～5.0μg/ml)and subtherapeutic group(<0.5μg/ml).Through single factor and multiple factor Logistic regression analysis,indeipendent predictive factors forvecz plasma concentration non-compliance were screened,and a column chart prediction model(NPM)was constructed.The performance of the model was evaluateding area under the receiver operating characteristic curve(AUC),Hosmer-Lemeshow(H-L)goodness-of-fit test,and decision curve analysis(DCA).Results Among 310 VRCZ-TDM measurements,71.61%(222/310)achieved target concentrations.Multivariate analysis showed that CYP2C19 intermediate metabolite,daily dose of cyclosporine A(CSA),daily dose of VRCZ,creatinine(Cr)>97 μmol/L,albumin(Alb)and C-reactive protein(CRP)were independent influencing factors for VRCZ blood drug concentration non-compliance(Wald χ2=4.046～13.221,all P＜0.05).The nomogram demonstrated excellent discrimination,calibration(H-L goodness of fit test χ2=2.663,P=0.954),and clinical utility with net benefit across 0.05～0.96 risk thresholds.Conclusion The nomogram incorporating CYP2C19 gene phenotype,daily CSA dosing,daily VRCZ dosing,Cr levels,Alb and CRP provides a validated tool for optimizing VRCZ therapy in allo-HSCT recipients,enabling precision dosing strategies.

Objective To identify determinants of subtherapeutic voriconazole(VRCZ)concentrations in allogeneic hematopoietic stem cell transplantation(allo-HSCT)recipients and to develop/validate a nomogram-based risk prediction model.Methods This study retrospectively analyzed 310 VRCZ therapeutic drug monitoring(TDM)measurements from allo-HSCT recipients at 310 patients who under went allo-HSCT surgery at Hebei Yanda Ludaopei Hospital from October 2022 to October 2024 and received VRCZ for the prevention and treatment of invasive fungal infections before transplantion were selected as the study subjects.Cases were stratified into target-concentration group(0.5～5.0μg/ml)and subtherapeutic group(<0.5μg/ml).Through single factor and multiple factor Logistic regression analysis,indeipendent predictive factors forvecz plasma concentration non-compliance were screened,and a column chart prediction model(NPM)was constructed.The performance of the model was evaluateding area under the receiver operating characteristic curve(AUC),Hosmer-Lemeshow(H-L)goodness-of-fit test,and decision curve analysis(DCA).Results Among 310 VRCZ-TDM measurements,71.61%(222/310)achieved target concentrations.Multivariate analysis showed that CYP2C19 intermediate metabolite,daily dose of cyclosporine A(CSA),daily dose of VRCZ,creatinine(Cr)>97 μmol/L,albumin(Alb)and C-reactive protein(CRP)were independent influencing factors for VRCZ blood drug concentration non-compliance(Wald χ2=4.046～13.221,all P＜0.05).The nomogram demonstrated excellent discrimination,calibration(H-L goodness of fit test χ2=2.663,P=0.954),and clinical utility with net benefit across 0.05～0.96 risk thresholds.Conclusion The nomogram incorporating CYP2C19 gene phenotype,daily CSA dosing,daily VRCZ dosing,Cr levels,Alb and CRP provides a validated tool for optimizing VRCZ therapy in allo-HSCT recipients,enabling precision dosing strategies.

Objective To systematically evaluate the published models in prediction of the risk of lung infections in elderly patients with hip fracture so as to provide a guidance for medical workers in selection or development of suitable risk prediction models.Methods Relevant studies were searched from databases including CNKI,Wanfang Data,VIP,SinoMed,PubMed,Web of Science,Cochrane Library,Embase and CINAHL,from the inception to 31st January,2024.Data were extracted from the selected literature and a bias assessment tool of risk predictive model was used to evaluate the risk of bias and applicability of the included literature.Results A total of 1,035 articles were retrieved,of which seven studies involving 13 predictive models were finally included after screening.The sample sizes ranged from 305 to 2,669 cases and lung infection rates ranged from 5.40%to 20.02%.The repeatedly reported predictors included age,gender,chronic obstructive pulmonary disease,hypoproteinaemia,American Society of Anesthesiologists(ASA)Physical Status Classification and white blood cell count.In the 13 models constructed,the reported area under the curve(AUC)of subjects'job characteristics ranged from 0.667 to 0.996.Five out of seven studies had good overall applicability,but all with high risk of bias.Conclusion The predictive models for lung infections in elderly patients with hip fracture are still in the stage of development.Although the predictive models show some predictive performance,however they are still deficient,and all studies have been found with a high risk in bias.

Objective:To investigate the application value of risk prediction model for acute kidney injury (AKI) after donation of cardiac death (DCD) liver transplantation based on machine learning algorithm.Methods:The retrospective cohort study was conducted. The clinicopathological data of 1 001 pairs of DCD liver transplant donors and recipients at five hospitals, including The First Affiliated Hospital of Zhejiang University School of Medicine et al, in the Chinese Liver Transplan-tation Registry from January 2015 to December 2023 were collected. Of the donors, there were 825 males and 176 females. Of the recipients, there were 806 males and 195 females, aged 52 (range, 18-75)years. There were 281 recipients included using oversampling technique, and all 1 282 recipients were divided to the training set of 897 recipients and the validation set of 385 recipients by a ratio of 7∶3 using computer-generated random numbers. Seven prediction models, including Random Forest (RF), Extreme Gradient Boosting (XGBoost), Support Vector Machine (SVM), Logistic Regression (LR), Decision Tree (DT), K-Nearest Neighbors (KNN), and Categorical Boosting (CatBoost), were constructed for AKI after liver transplantation based on machine learning algorithm. Observation indicators: (1) comparison of clinicopathological characteristics between recipients with and without AKI and donors; (2) follow-up and survival of recipients with and without AKI; (3) construction and validation of nomogram prediction model of AKI after liver transplantation; (4) construction and validation of machine learning prediction model of AKI after liver transplantation. Comparison of measurement data with normal distribution between groups was conducted using the independent sample t test. Comparison of measurement data with skewed distribution between groups was conducted using the Mann-Whitney U test, and comparison among groups was conducted using the Kruskal-Wallis H test. Comparison of count data between groups was conducted using the chi-square test or corrected chi-square test. Kaplan-Meier method was used to calculate survival rates and plot survival curves. Logistic regression model was performed for univariate and multivariate analyses. The receiver operating characteristic (ROC) curve was plotted to calculate area under curve (AUC) and 95% confidence interval ( CI). The performance of prediction model was evaluated using DeLong test, accuracy, sensitivity, specificity. The calibration curve was plotted to evaluate the performance of predicted probability and actual probability. The interpretability analysis of machine learning algorithm and SHapley Additive exPlanations was used to explain the model decision separately. Results:(1) Comparison of clinicopathological characteristics between recipients with and without AKI and donors. Of 1 001 recipients, there were 360 cases with AKI and 641 cases without AKI after liver transplantation. There were significant differences in body mass index (BMI), hepatic encepha-lopathy, hepatitis B surfact antigen (HBsAg), hepatorenal syndrome (HRS) and donor diabetes, donor blood urea nitrogen, donor alanine aminotransferase, donor aspartate aminotransferase, mass of graft, volume of blood loss during liver transplantation, warm ischema time of donor liver, and operation time between recipients with and without AKI ( Z=-4.337, χ2=9.751, 9.088, H=11.142, χ2=5.286, Z=-3.360, -2.539, -3.084, -1.730, -3.497, -1.996, -2.644, P<0.05). (2) Follow-up and survival of recipients with and without AKI. All the 1 001 recipients received follow-up. The recipients with AKI after liver transplantation were followed up for 18.6(range, 0-102.3)months, and recipients without AKI after liver transplantation were followed up for 31.9(range, 0.1-105.5)months. The 1-, 3-, and 5-year overall survival rates were 72.1%, 63.5%, and 59.3% of recipients with AKI, versus 86.7%, 76.7%, and 72.5% of recipients without AKI, respectively, showing a significant difference in overall survival between them ( χ2=26.028, P<0.05). (3) Construction and validation of nomogram predic-tion model of AKI after liver transplantation. Results of multivariate analysis showed that recipient BMI, recipient creatinine, recipient HBsAg, recipient HRS, donor blood urea nitrogen, donor crea-tinine, anhepatic phase and volume of blood loss during liver transplantation were independent risk factors for AKI of recipients after liver transplantation ( odds ratio=1.113, 0.998, 0.605, 1.580, 1.047, 0.998, 1.006, 1.157, 95% CI as 1.070-1.157, 0.996-1.000, 0.450-0.812, 1.021-2.070, 1.021-1.074, 0.996-0.999, 1.000-1.012, 1.045-1.281, P<0.05). The nomogram prediction model of AKI after liver transplantation was constructed based on the results of multivariate analysis. Results of ROC curve showed that the AUC of 0.666 (95% CI as 0.637-0.696). (4) Construction and validation of machine learning prediction model of AKI after liver transplantation. Based on the Lasso regression analysis, seven machine learning algorithm prediction models, including RF, XGBoost, SVM, LR, DT, KNN, and CatBoost, were constructed, with ROC curves of the validation set plotted. The AUC of above models were 0.863, 0.841, 0.721, 0.637, 0.620, 0.708, 0.731, accuracies were 0.764, 0.782, 0.701, 0.592, 0.605, 0.605, 0.681, sensitivities were 0.764, 0.789, 0.719, 0.588, 0.694, 0.694, 0.704, specificities were 0.763, 0.774, 0.683, 0.597, 0.511, 0.511, 0.656, respectively. Delong test showed that the RF model with the highest AUC of 0.863(95% CI as 0.828-0.899). Calibration curve analysis showed the predicted probability closest to the actual probability of RF model, indicating the model with a good validation value. Further sorting of SHAP of different clinical factors based on RF model showed that recipient BMI, donor blood urea nitrogen, volume of blood loss during liver transplantation, donor age had large effects on the output outcomes. Conclusion:The nomogram prediction model and seven machine learning algorithm prediction models for AKI after DCD liver transplantation are constructed, and the RF model based on machine learning has a better predictive performance.

[Objective]To examine the botanical origin of the exotic medicinal materia Alebo firstly contained in Bencao Shiyi.[Methods]Based on the records of Haiyao Bencao,Compendium of Materia Medica,etc.,through ancient and modern literature verification and comparative botanical studies,combined with fieldwork in the field,from the perspectives of plant morphology,habitat distribution,herb name,nature and flavor,efficacy,the Alebo was examined.[Results]Comparative study revealed that the Alebo(Brahman saponaria)of the Bencao Shiyi matches the morphological characteristics of the Leguminosae Tamarindus indica L.(T.indica).T.indica are native to the Indian Peninsula and the Malay Archipelago,matching the habitat and name of the Brahmin saponaria.The two properties of similar nature and flavor,efficacy and mouth taste of T.indica peel and seeds are in line with the Brahmin Saponaria of the"taste".The Compendium of Materia Medica blackmail variant of Abole(Persian saponaria)is morphologically consistent with the Leguminosae Cassia fistula L.(C.fistula),and also bears some of the characteristics of the Kigelia africana(Lam.)Benth.(K.africana).In conjunction with further research,the habitat,name,and nature and flavor,efficacy effects of the C.fistula are consistent with the characteristics of Abole.Compendium of Materia Medica appendix Guihai Zhi tamarind,Sterculiaceae Sterculia nobilis Smith(S.nobilis)in line with the"nine layers of skin fruit"one.S.nobilis and Sterculia lanceolata Cav.(S.lanceolata)are consistent with tamarind in morphology,habitat,and nature and flavor,efficacy.[Conclusion]It is suggested that in the revision of the Dictionary of Chinese Medicine and other books,the origin of"Brahmin saponaria"should be changed to T.indica of the Leguminosae,and that in the revision of the Chinese Materia Medica Uyghur Medicine Volume and other books,the origin of"tamarind"should be changed to S.nobilis and S.lanceolata of the Sterculiaceae.In order to correct the origin,renew the history of medicine,inherit the essence.

Objective:To explore the efficacy of the features of ultrasound imaging omics at the area peripheral thyroid nodule in predicting cervical lymph node metastasis(LNM)of papillary thyroid carcinoma(PTC),and construct a prediction model based on the features of imaging omics and to verify its performance.Methods:A total of 237 PTC patients who admitted to Shaanxi Provincial Hospital of Chinese Medicine from March 2021 to June 2024 and were confirmed by pathology were retrospectively collected.They were divided into the training set(166 cases)and the validation set(71 cases)as a ratio of 7 to 3.According to the postoperatively pathological results,237 patients were divided into the metastasis group(108 cases)and the non-metastasis group(129 cases).The clinical data and conventional ultrasound characteristic information of all patients were collected,and a feature model of imaging omics was constructed through quantitative extracting and screening the features of ultrasound imaging omics within nodules and peripheral nodules,and utilizing machine learning classifier.Then,the feature score(Rad-Score)of image omics was obtained.The Rad-Score values within and peripheral nodules,and the Rad-Score values peripheral nodules of metastasis group and non-metastasis group were compared.In training set,the independent risk factors of affecting neck LNM were analyzed,and a clinical-ultrasonic model was constructed,which was combined with Rad-Score to construct a joint model based on the features of imaging omics peripheral nodules.The receiver operating characteristic(ROC)curve was used to analyze and compare its predictive efficacy.The nomogram of the joint model was constructed,and then,the calibration and fitting degrees of the nomogram were assessed by calibration curve and Hosmer-Lemeshow test.Result:In training set,6 features of imaging omics within nodules and 11 features of imaging omics peripheral nodules were respectively extracted and screened out through the least absolute shrinkage and selection operator(LASSO)algorithm.In the training set and validation set,the Rad-Scores peripheral the nodules in the metastasis group were respectively(7.43±0.45)points and(7.19±0.51)points,which were significantly higher than(3.25±0.28)points and(3.51±0.32)points peripheral the nodules in the non-metastasis group,and the differences were statistically significant(t=72.708,61.222,P<0.05).The results of factor analysis showed that age,capsule invasion,microcalcification,ultrasound-indicated lymph node positivity and Rad-Score around nodules were independent risk factors of affecting cervical LNM of PTC patients(OR=0.592,2.983,3.593,4.424,2.575,P<0.05).The ROC curve showed that the area under curve(AUC)values of the ROC curve of joint model in training set and validation set were respectively 0.861 and 0.872 in predicting LNM,respectively,which were superior to 0.759 and 0.783 of the clinical-ultrasound model.Conclusion:In both the training set and the validation set,the nomogram of joint model has favorable calibration and fitting in predicting cervical LNM of PTC patients.The construction of clinical model based on the features of ultrasound imaging omics peripheral nodules has a favorable efficacy in predicting the status of cervical lymph node of PTC patients before surgery,which is expected to be an effective tool of individual prediction for LNM.