Abstract:Introduction: To investigate the risk factors affecting patients with placenta previa (PP) and to construct an effective prediction model for the severity of PAS in PP. Methods: A total of 240 pregnant women with PP were enrolled in this study. An MRI+Ultrasound-based model was developed to classify patients into placental implantation and non-placental implantation groups. Multivariate nomograms were created based on imaging features. The model was evaluated using Receiver Operating Characteristic (ROC) curve analysis. The predictive accuracy of the nomogram was assessed through calibration plots and decision curve analysis. Results: The MRI+Ultrasound-based prediction model demonstrated favorable discrimination between the placental implantation and non-placental implantation groups. The calibration curve exhibited agreement between the estimated and actual probability of placental implantation. Additionally, decision curve analysis indicated a high clinical benefit across a wide range of probability thresholds. The Area under the ROC curve (AUC) was 0.911 (95 % CI: 0.76-0.947), with a sensitivity of 88.40 % and specificity of 88.10 %. Conclusion: The MRI+Ultrasound-based prediction model could be a valuable tool for preoperative prediction of the percentage of implantation. Our study enables obstetricians to conduct more adequate preoperative evaluations.

A prediction model based on MRI and ultrasound to predict the risk of PAS in patient with placenta previa