Diagnostic Value of Magnetic Resonance Imaging(MRI)T2WI Sequence Radiomics Model for Special Site Ectopic Pregnancy
Objective To construct a radiomics model based on MRI T2WI sequence images and explore its diagnostic value for special site ectopic pregnancy.Methods A total of 258 ectopic pregnancy patients who underwent treatment in our hospital from February 2018 to February 2023 were selected.The patients were divided into a training set(n=186)and a validation set(n=72)in a ratio of 2:1.Based on histopathological results,the patients from training set were divided into special site ectopic pregnancy group(n=102)and non-special site ectopic pregnancy group(n=84).Multivariable logistic regression was performed to identify independent factors associated with special site ectopic pregnancy.Univariate analysis and Least Absolute Shrinkage and Selection Operator(LASSO)regression were used to reduce the dimensionality of the radiomics features obtained from MRI T2WI sequence images.A combined model of radiomics and clinical features was constructed,and the predictive ability of the combined model model was evaluated using receiver operating characteristic(ROC)curve,Hosmer-Lemeshow test,and calibration curve.The clinical utility of the model was assessed using decision curve analysis(DCA).Results The MRI has a diagnostic accuracy of 79.07％,specificity of 80.29％,sensitivity of 77.69％,positive predictive value(PPV)of 77.69％,and negative predictive value(NPV)of 80.29％for diagnosing ectopic pregnancy in ectopic pregnancy patients.Compared to non-special sites ectopic pregnancies,there were significant differences in the number of induced abortions,ectopic pregnancy history,β-hCG ratio,endometrial thickness,and MRI diagnosis of special site ectopic pregnancies(P＜0.05).Multivariate Logistic regression analysis showed that low β-hCG ratio,high frequency of induced abortions,history of ectopic pregnancy,thin endometrium,and positive MRI diagnosis of special site ectopic pregnancy were independent influencing factors for special site ectopic pregnancy compared to non-special site ectopic pregnancy.After dimension reduction through one-way ANOVA and LASSO regression,nine representative radiomic features and clinical features were obtained to form a joint model.The AUC values for the joint model,the radiomic score model,and the clinical prediction model were 0.837,0.741,and 0.717 respectively in the training set,and 0.816,0.752,and 0.705 respectively in the validation set.The joint model showed the highest AUC,indicating superior predictive performance(P＜0.05).The Hosmer-Lemeshow test showed no statistical difference between the predicted and observed data in the combined model(P=0.146,0.102).The calibration curve showed good agreement between predicted and observed values in the validation and training sets.The decision curve analysis(DCA)curve showed that the combined model and radiomics score model had a higher predictive value within a large range of risk thresholds.Cancteion The MRI T2WI sequence radiomics model has high accuracy in predicting special site ectopic pregnancy and is of great clinical value.
Magnetic Resonance imagingT2WI sequenceRadiomicsSpecial Site Ectopic PregnancyLASSO
NETL
NSTL
万方数据
