ObjectiveTo develop a predictive model and application for spontaneous passage of common bile duct stones using automated machine learning algorithms given the complexity of treatment decision-making for patients with common bile duct stones， and to reduce unnecessary endoscopic retrograde cholangiopancreatography （ERCP） procedures. MethodsA retrospective analysis was performed for the data of 835 patients who were scheduled for ERCP after a confirmed diagnosis of common bile duct stones based on imaging techniques in Changshu First People’s Hospital （dataset 1） and Changshu Traditional Chinese Medicine Hospital （dataset 2）. The dataset 1 was used for the training and internal validation of the machine learning model and the development of an application， and the dataset 2 was used for external testing. A total of 22 potential predictive variables were included for the establishment and internal validation of the LASSO regression model and various automated machine learning models. The area under the receiver operating characteristic curve （AUC）， sensitivity， specificity， and accuracy were used to assess the performance of models and identify the best model. Feature importance plots， force plots， and SHAP plots were used to interpret the model. The Python Dash library and the best model were used to develop a web application， and external testing was conducted using the dataset 2. The Kolmogorov-Smirnov test was used to examine whether the data were normally distributed， and the Mann-Whitney U test was used for comparison between two groups， while the chi-square test or the Fisher’s exact test was used for comparison of categorical data between groups. ResultsAmong the 835 patients included in the study， 152 （18.20%） experienced spontaneous stone passage. The LASSO model achieved an AUC of 0.875 in the training set （n=588） and 0.864 in the validation set （n=171）， and the top five predictive factors in terms of importance were solitary common bile duct stones， non-dilated common bile duct， diameter of common bile duct stones， a reduction in serum alkaline phosphatase （ALP）， and a reduction in gamma-glutamyl transpeptidase （GGT）. A total of 55 models were established using automated machine learning， among which the gradient boosting machine （GBM） model had the best performance， with an AUC of 0.891 （95% confidence interval： 0.859‍ ‍—‍ ‍0.927）， outperforming the extreme randomized tree mode， the deep learning model， the generalized linear model， and the distributed random forest model. The GBM model had an accuracy of 0.855， a sensitivity of 0.846， and a specificity of 0.857 in the test set （n=76）. The variable importance analysis showed that five factors had important influence on the prediction of spontaneous stone passage， i.e.， were solitary common bile duct stones， non-dilated common bile duct， a stone diameter of <8 mm， a reduction in serum ALP， and a reduction in GGT. The SHAP analysis of the GBM model showed a significant increase in the probability of spontaneous stone passage in patients with solitary common bile duct stones， non-dilated common bile duct， a stone diameter of <8 mm， and a reduction in serum ALP or GGT. ConclusionThe GBM model and application developed using automated machine learning algorithms exhibit excellent predictive performance and user-friendliness in predicting spontaneous stone passage in patients with common bile duct stones. This application can help avoid unnecessary ERCP procedures， thereby reducing surgical risks and healthcare costs.

Objective:To construct and validate a predictive model and application program for stone recurrence after endoscopic retrograde cholangiopancreatography (ERCP) based on machine learning algorithms in patients with common bile duct stones (CBDS).Methods:A multicenter retrospective cohort study was conducted, 862 CBDS patients underwent ERCP from June 2020 to September 2023 in Changshu First People′s Hospital (data set 1, 759 cases, including a training set of 588 cases and a validation set of 171 cases) and Changshu Hospital of Traditional Chinese Medicine (data set 2, 103 cases, used as a test set). The demographics, medical history, ERCP procedural records and laboratory indices were collected. All patients were followed up for 1 year, and the stone recurrence was recorded. In training set, the feature selection was conducted by the least absolute shrinkage and selection operator (LASSO) algorithm, and a conventional Logistic regression model was constructed based on selected features. The 3 machine learning algorithms (gradient boosting machine model, extreme gradient boosting model and random forest model) and a conventional Logistic regression model (LASSO model) were trained to fit predictive models. The model performance was assessed by area under curve (AUC) of receiver operating characteristic curve. The model interpretability was analyzed by feature importance evaluation, Shapley additive explanations (SHAP) and force plots. The best-performing model was deployed as an online application by Streamlit framework (V1.36.0).Results:Among the 862 patients, 158 patients (18.33%) developed stone recurrence after ERCP. There were no statistical difference in demographics, medical history, ERCP procedural records and laboratory indices between training set and a validation set ( P>0.05). LASSO regression analysis result showed that 6 key variables (in descending order of significance: endoscopic sphincterotomy, common bile duct angulation, stone diameter, stone count, common bile duct diameter, and periampullary diverticulum) influencing stone recurrence. ROC curve analysis result showed that the random forest model exhibited the highest predictive performance (it had the largest AUC of 0.900). SHAP analysis result showed that common bile duct angulation, common bile duct diameter, stone diameter, endoscopic sphincterotomy and stone count were the top 5 contributing factors in the random forest model. Using Python, the random forest model was implemented into a Streamlit-based application with a user-friendly visual interface, providing predictive outcomes, confidence levels, SHAP force diagram and health recommendations. In the test set, the application program achieved an accuracy of 84.5% (87/103), sensitivity of 82.6% (19/23), and specificity of 85.0% (68/80). SHAP plots and force diagram intuitively illustrated the impact of key features on stone recurrence prediction, offering a clear visualization of each variable′s role within the model. Conclusions:The predictive model and application program based on the random forest machine learning algorithms demonstrate excellent predictive performance and practical usability in predicting stone recurrence after ERCP in patients with CBDS.

Objective To develop a combined model integrating radiomics and 3D deep learning features for improving the predictive efficacy of overall survival in non-small cell lung cancer(NSCLC)patients undergoing radiotherapy,thereby providing a foundation for optimizing individualized radiotherapy strategies.Methods A retrospective analysis was conducted on 522 NSCLC patients from 3 centers.Radiomics features were extracted from the tumor region of interest on radiotherapy planning CT scans,and a 3D-SE-ResNet was constructed to extract deep learning features.Following feature extraction,features were selected via univariate Cox analysis and Lasso-Cox regression,and a combined model was established by fusing the two feature types through principal component analysis.The discriminative ability of the model was evaluated using the concordance index(C-index)and the area under the receiver operating characteristic curve(AUC),while the risk stratification efficacy was verified by Kaplan-Meier survival analysis.Results The predictive performance of deep learning features was significantly superior to that of radiomics features(C-index:0.73 vs 0.65).The combined model achieved the highest predictive performance in the training set,internal test set,and external test set(C-index:0.74,0.69,0.72 respectively),with higher AUC values for predicting 1-year,2-year,and 3-year OS than either single model.Kaplan-Meier analysis showed significant differences in survival between the high-and low-risk groups(Log-rank test,P<0.001),and calibration curves indicated good consistency between predicted and actual survival outcomes.Conclusion The combined model integrating radiomics and 3D deep learning features can accurately predict survival outcomes in NSCLC patients undergoing radiotherapy.The multi-center validation results support its potential application in prognosis stratification for individualized radiotherapy.

Objective To explore the risk factors for postoperative bleeding after endoscopic resection of gastric gastrointestinal stromal tumor(gGIST)and to construct prediction models using 4 different machine learning algorithms for accurately predicting postoperative bleeding.Methods The clinical data of gGIST patients were collected,and the patients were randomly divided into a training cohort(n=502)and a validation cohort(n=130)at an 8:2 ratio.Synthetic minority over-sampling technique-nominal continuous was used for oversampling in the training cohort.Four prediction models were constructed using gradient boost machine(GBM),deep learning,generalized linear model and distributed random forest,separately;and in addition,the least absolute shrinkage and selection operator was used to screen variables and construct a traditional Logistic regression model.Model performance was evaluated by calculating the area under the receiver operating characteristic curve(AUC),sensitivity,specificity,accuracy,positive predictive value and negative predictive value.Interpretability analyses,including feature importance,SHapley additive exPlanation and force plot,were performed on the optimal model,and a practically applicable web application was developed.Results Among 632 patients,78(12.3%)experienced postoperative bleeding.In the validation cohort,GBM model performed best among 5 prediction models,with an AUC value of 0.889 and a 95%CI of 0.829-0.948,superior to the other 4 models.Variable importance analysis identified surgeon experience,operation time,intraoperative hemorrhage,tumor size as the factors affecting postoperative bleeding prediction.The SHapley additive exPlanation plot and force plot showed the distribution characteristics of variables in the binary classification prediction results and the effect of each variable on the prediction results.Conclusion GBM model has high predictive value for postoperative bleeding following endoscopic resection of gGIST,and the construction of the web application facilitates its clinical use.

Objective:To explore the relationship between programmed death-1(PD-1)inhibitors and cancer-related fatigue(CRF)in patients with malignant melanoma,and to identify associated influencing factors.Methods:A total of 100 patients with malignant melanoma treated at Jinan People's Hospital between April 2019 and April 2024 were included as study subjects.The Chinese version of the Piper Fatigue Scale was used to evaluate patients'fatigue levels within three months before and after the initiation of PD-1 inhibitor therapy.Results:There was a significant difference in CRF score before and after PD-1 inhibitor treatment(P<0.001).Univariate analysis showed no significant association between fatigue severity and factors such as gender,smoking history,tumor site,or PD-1 inhibitor type(all P>0.05).However,age,tumor stage,anemia,leukopenia,secondary hypothyroidism,secondary adrenal insufficiency(AI),and secondary adrenocorticotropic hormone deficiency(P<0.05 or P<0.01 or P<0.001)were significantly associated with CRF.Multivariate regression analysis identified secondary hypothyroidism,secondary AI,anemia,and leukopenia as independent risk factors for severe CRF in patients with malignant melanoma(all P<0.05).Conclusion:Adverse reactions of PD-1 inhibitors,including secondary hypothyroidism,secondary AI,anemia,and leukopenia,are independent risk factors for CRF in patients with malignant melanoma.

Objective To develop a combined model integrating radiomics and 3D deep learning features for improving the predictive efficacy of overall survival in non-small cell lung cancer(NSCLC)patients undergoing radiotherapy,thereby providing a foundation for optimizing individualized radiotherapy strategies.Methods A retrospective analysis was conducted on 522 NSCLC patients from 3 centers.Radiomics features were extracted from the tumor region of interest on radiotherapy planning CT scans,and a 3D-SE-ResNet was constructed to extract deep learning features.Following feature extraction,features were selected via univariate Cox analysis and Lasso-Cox regression,and a combined model was established by fusing the two feature types through principal component analysis.The discriminative ability of the model was evaluated using the concordance index(C-index)and the area under the receiver operating characteristic curve(AUC),while the risk stratification efficacy was verified by Kaplan-Meier survival analysis.Results The predictive performance of deep learning features was significantly superior to that of radiomics features(C-index:0.73 vs 0.65).The combined model achieved the highest predictive performance in the training set,internal test set,and external test set(C-index:0.74,0.69,0.72 respectively),with higher AUC values for predicting 1-year,2-year,and 3-year OS than either single model.Kaplan-Meier analysis showed significant differences in survival between the high-and low-risk groups(Log-rank test,P<0.001),and calibration curves indicated good consistency between predicted and actual survival outcomes.Conclusion The combined model integrating radiomics and 3D deep learning features can accurately predict survival outcomes in NSCLC patients undergoing radiotherapy.The multi-center validation results support its potential application in prognosis stratification for individualized radiotherapy.

Objective:To construct and validate a predictive model and application program for stone recurrence after endoscopic retrograde cholangiopancreatography (ERCP) based on machine learning algorithms in patients with common bile duct stones (CBDS).Methods:A multicenter retrospective cohort study was conducted, 862 CBDS patients underwent ERCP from June 2020 to September 2023 in Changshu First People′s Hospital (data set 1, 759 cases, including a training set of 588 cases and a validation set of 171 cases) and Changshu Hospital of Traditional Chinese Medicine (data set 2, 103 cases, used as a test set). The demographics, medical history, ERCP procedural records and laboratory indices were collected. All patients were followed up for 1 year, and the stone recurrence was recorded. In training set, the feature selection was conducted by the least absolute shrinkage and selection operator (LASSO) algorithm, and a conventional Logistic regression model was constructed based on selected features. The 3 machine learning algorithms (gradient boosting machine model, extreme gradient boosting model and random forest model) and a conventional Logistic regression model (LASSO model) were trained to fit predictive models. The model performance was assessed by area under curve (AUC) of receiver operating characteristic curve. The model interpretability was analyzed by feature importance evaluation, Shapley additive explanations (SHAP) and force plots. The best-performing model was deployed as an online application by Streamlit framework (V1.36.0).Results:Among the 862 patients, 158 patients (18.33%) developed stone recurrence after ERCP. There were no statistical difference in demographics, medical history, ERCP procedural records and laboratory indices between training set and a validation set ( P>0.05). LASSO regression analysis result showed that 6 key variables (in descending order of significance: endoscopic sphincterotomy, common bile duct angulation, stone diameter, stone count, common bile duct diameter, and periampullary diverticulum) influencing stone recurrence. ROC curve analysis result showed that the random forest model exhibited the highest predictive performance (it had the largest AUC of 0.900). SHAP analysis result showed that common bile duct angulation, common bile duct diameter, stone diameter, endoscopic sphincterotomy and stone count were the top 5 contributing factors in the random forest model. Using Python, the random forest model was implemented into a Streamlit-based application with a user-friendly visual interface, providing predictive outcomes, confidence levels, SHAP force diagram and health recommendations. In the test set, the application program achieved an accuracy of 84.5% (87/103), sensitivity of 82.6% (19/23), and specificity of 85.0% (68/80). SHAP plots and force diagram intuitively illustrated the impact of key features on stone recurrence prediction, offering a clear visualization of each variable′s role within the model. Conclusions:The predictive model and application program based on the random forest machine learning algorithms demonstrate excellent predictive performance and practical usability in predicting stone recurrence after ERCP in patients with CBDS.

Objective To explore the risk factors for postoperative bleeding after endoscopic resection of gastric gastrointestinal stromal tumor(gGIST)and to construct prediction models using 4 different machine learning algorithms for accurately predicting postoperative bleeding.Methods The clinical data of gGIST patients were collected,and the patients were randomly divided into a training cohort(n=502)and a validation cohort(n=130)at an 8:2 ratio.Synthetic minority over-sampling technique-nominal continuous was used for oversampling in the training cohort.Four prediction models were constructed using gradient boost machine(GBM),deep learning,generalized linear model and distributed random forest,separately;and in addition,the least absolute shrinkage and selection operator was used to screen variables and construct a traditional Logistic regression model.Model performance was evaluated by calculating the area under the receiver operating characteristic curve(AUC),sensitivity,specificity,accuracy,positive predictive value and negative predictive value.Interpretability analyses,including feature importance,SHapley additive exPlanation and force plot,were performed on the optimal model,and a practically applicable web application was developed.Results Among 632 patients,78(12.3%)experienced postoperative bleeding.In the validation cohort,GBM model performed best among 5 prediction models,with an AUC value of 0.889 and a 95%CI of 0.829-0.948,superior to the other 4 models.Variable importance analysis identified surgeon experience,operation time,intraoperative hemorrhage,tumor size as the factors affecting postoperative bleeding prediction.The SHapley additive exPlanation plot and force plot showed the distribution characteristics of variables in the binary classification prediction results and the effect of each variable on the prediction results.Conclusion GBM model has high predictive value for postoperative bleeding following endoscopic resection of gGIST,and the construction of the web application facilitates its clinical use.

Objective To propose a deep learning neural network approach for transforming megavoltage computed tomography(MVCT)images of cervical cancer into pseudo kilovoltage computed tomography(kVCT)images with high signal-to-noise ratio and contrast-to-noise ratio,thus providing three-dimensional anatomical images and localization information required for adaptive radiotherapy of cervical cancer,and guiding the accelerator to achieve precise treatment.Methods The MVCT and kVCT images of 54 patients treated with cervical cancer radiotherapy were collected,with 44 cases randomly selected as the training set,and the remaining 10 cases as the test set.A cyclic generative adversarial network with gating mechanism and multi-channel data input was used to synthesize pseudo-kVCT images from MVCT images.The network training results were evaluated with imaging quality evaluation parameters,such as mean absolute error(MAE),peak signal-to-noise ratio(PSNR),and structural similarity index(SSIM).Results The MAE,PSNR,and SSIM of MVCT imagesvspseudo-kVCT(5:5)images were(24.9±0.7)HUvs(17.8±0.3)HU,(29.8±0.2)dBvs(30.7±0.2)dB,and 0.841±0.007 vs 0.898±0.003,respectively.Conclusion The generated pseudo-kVCT images have advantages in noise reduction and contrast enhancement,and can reduce the need for additional MV-kVCT electron density calibration in dose calculations.The dose calculation ability of pseudo-kVCT is comparable to that of MVCT,providing a possibility for the application of pseudo-kVCT images in image-guided adaptive radiotherapy.

【Objective】 To systematically evaluate the efficacy and safety of targeted drugs in the treatment of metastatic non-clear cell renal cell carcinoma (nccRCC) and to provide guidance for clinical treatment. 【Methods】 All observational studies and randomized controlled trials (RCTs) of nccRCC treated with targeted drugs were retrieved from the PubMed, Embase, the Cochrane Library and Web of Science. Three independent investigators screened the literature, extracted data and evaluated the quality of literature. The RCTs were evaluated using the Cochrane Handbook. One research with insufficient outcome data (follow-up bias) was assessed as high risk, and the other studies showed low or uncertain risk. The non-RCTs were evaluated with the JBI Quality Assessment Tool, and all studies displayed a low risk of bias. The data were analyzed with Stata 17.0 software. 【Results】 A total of 16 studies involving 989 patients were included. Meta-analysis showed that the objective response rate (ORR) was 12.6% (95%CI:8.1%-17.9%), the total disease control rate (DCR) was 65.3% (95%CI:58.3%-72.1%), the total median progression-free survival (PFS) was 5.80 (95%CI:4.69-6.91) months, and the median overall survival (OS) was 15.93 (95%CI:12.17-19.68) months. In subgroup analysis, the total ORR of patients with metastatic nccRCC treated with sunitinib and cabozantinib were 11.7% (95%CI:6.5%-18.0%) and 17.2% (95%CI:8.4%-28.2%), respectively. The total ORR of patients with papillary renal cell carcinoma was 9.1% (95%CI:2.4%-18.9%). 【Conclusion】 Targeted drugs have a significant effect on patients with metastatic nccRCC, but adverse reactions may occur. Targeted drugs have poor effects on metastatic papillary renal cell carcinoma, and cabozantinib may have greater survival benefits.