Preliminary Study On The Application Of Nomogram Based On Gd-EOB-DTPA Enhanced MRI Radiomics And Clinical Characteristics To Predict The Grading Of Esophageal Varices

Purpose: To explore the value of nomogram based on gadolinium-ethoxybenzyldiethylenetriamine pentaacetic acid(Gd-EOB-DTPA)enhanced MRI-based radiomics combined with clinical characteristics to predict the grading of esophageal varices(EV)in patients with liver cirrhosis.Methods: This study retrospectively included 190 EV patients with liver cirrhosis who underwent gastroscopy and Gd-EOB-DTPA enhancement examination in the first Hospital of Lanzhou University from January 2020 to January 2023,and their clinical data,serological results and MR imaging data were collected.The cases were randomly divided into a training group and a validation group in a 7:3 ratio.The 3D-Slicer software was used to draw the regions of interest(ROI)on the portal venous phase and hepatobiliary phase images,and the corresponding radiomics features were extracted.Pearson correlation coefficient(PCC),principal component analysis(PCA),analysis of variance(ANOVA)and recursive feature elimination(RFE)were used for feature dimension reduction and selection.Logistic regression(LR),support vector machine(SVM)and linear discriminant analysis(LDA)were used to build radiomics models.The nomogram was constructed by combining the clinical independent risk factors and the radiomics score(Radscore)of the best radiomics model.Receiver operating characteristic(ROC)curves were used to evaluate the predictive performance of the nomogram.Calibration curves were used to assess the accuracy and reliability of the predictive model.Decision curve analysis(DCA)was conducted to evaluate the clinical applicability of the model.Results:1.Univariate analysis showed that among the clinical serological factors,there were statistically significant differences between groups in white blood cells,platelets,prothrombin time,alanine aminotransferase,international normalized ratio(INR)and hemoglobin(HGB)(P<0.05).Multivariate Logistic regression analysis showed that only hemoglobin(HGB)and international normalized ratio(INR)were independent risk factors associated with EV grading(P<0.05).2.The radiomics model based on the combination of portal venous phase and hepatobiliary phase sequences demonstrated good diagnostic performance.In the training group,the area under the curve(AUC),accuracy,sensitivity,and specificity were 0.830,0.807,0.737,and 0.867,respectively.In the validation group,the corresponding values were 0.727,0.714,0.600,and 0.895,respectively.3.The nomogram model constructed based on independent clinical risk factors and radiomics score outperformed the individual clinical model and radiomics model alone.The area under the curve(AUC)for the training group and validation group were0.882(95% CI: 0.808-0.957)and 0.872(95% CI: 0.750-0.993),respectively.The calibration curve demonstrated good consistency between the predicted EV grading and the actual results.The decision curve analysis(DCA)curve indicated that the nomogram model had higher clinical applicability compared to the clinical model.Conclusion: HGB and INR were identified as independent clinical risk factors for predicting the grading of EV in patients with liver cirrhosis.The nomogram based on portal venous phase and hepatobiliary phase radiomics characteristics combined with clinical serological characteristics exhibits good diagnostic performance in predicting the grading of EV.It holds potential for non-invasive diagnosis of EV grading in liver cirrhosis and has certain potential in the noninvasive diagnosis the grading of EV in cirrhosis.