生殖医学杂志

目的 构建临床因素联合磁共振成像(MRI)预测胎盘植入性疾病(PAS)不良妊娠结局的列线图模型，并评估其准确性和临床应用价值。方法 回顾性分析2021年7月至2023年12月在常州市妇幼保健院接受诊治的PAS患者的临床资料。所有患者均行MRI检查，记录影像学特征。采用Logistic回归分析筛选不良妊娠结局的独立预测因素，基于这些因素构建列线图模型，并通过受试者工作特征(ROC)曲线和校准曲线评估模型的预测性能。结果 临床因素方面，PAS不良妊娠结局组患者术前血红蛋白水平显著低于无不良妊娠结局组(P<0.05),完全性前置胎盘比例显著高于无不良妊娠结局组(P<0.05)。MRI检查方面，不良妊娠结局组胎盘内T2WI低信号带、胎盘内异常增生血管、子宫肌层变薄、子宫外突和胎盘突入膀胱征比例显著高于未发生不良妊娠结局组(P<0.05)。Logistic回归分析显示，术前低血红蛋白水平、合并完全性前置胎盘、胎盘内T2W1低信号带、胎盘内异常增生血管、子宫肌层变薄、子宫外突、胎盘突入膀胱征象是PAS患者发生不良妊娠结局的独立危险因素(P<0.05)。基于这些独立预测因素构建的列线图模型ROC的敏感度为85.7%、特异度为83.0%,曲线下面积(AUC)为0.909,预测性能良好。校准曲线显示模型预测值与实际观测值一致性良好。结论 本研究构建的临床因素联合MRI的列线图模型能够较好地预测PAS不良妊娠结局，具有一定的临床应用价值，可为临床医师的诊断治疗提供依据和参考。

Objective:To construct a nomogram model of clinical factors combined with magnetic resonance imaging(MRI) for predicting adverse pregnancy outcomes in the patients with placenta accreta spectrum(PAS) disorders, and to evaluate its accuracy and clinical apply value.Methods:A retrospective analysis of clinical data of PAS patients treated in Changzhou Maternal and Child Health Hospital from July 2021 to December 2023 was conducted. All patients underwent MRI examination, and imaging characteristics were recorded. Logistic regression analysis was used to identify independent predictors of adverse pregnancy outcomes, which were then used to construct a nomogram model. The predictive performance of the model was evaluated using receiver operating characteristic(ROC) curves and calibration curves.Results:In terms of clinical factors, the preoperative hemoglobin levels were lower in adverse pregnancy outcome group than that in non-adverse pregnancy outcome group(P<0.05). The proportion of complete placenta previa was higher in adverse pregnancy outcome group when compared with non-adverse pregnancy outcome group(P<0.05). In terms of MRI findings, the adverse pregnancy outcome group had a higher proportion of low-signal bands on T2-weighted images(T2WI),abnormal proliferative vessels within the placenta, thinning of the myometrium, uterine bulging, and the bladder invasion by the placenta(P<0.05). The results of logistic regression analysis showed that preoperative low hemoglobin levels, complete placenta previa, low-signal bands on T2WI,abnormal proliferative vessels within the placenta, thinning of the myometrium, uterine bulging, and the bladder invasion by the placenta were independent risk factors for adverse pregnancy outcomes(P<0.05). The nomogram model based on these independent predictors showed a sensitivity of 85.7%,a specificity of 83.0%,and an area under the curve(AUC) of 0.909,indicating good predictive performance. The calibration curve demonstrated good consistency between predicted and actual observed values.Conclusions:The nomogram model constructed in this study, which combines clinical factors with MRI,can effectively predict adverse pregnancy outcomes in PAS. It holds certain clinical value and can provide a basis and reference for clinicians' diagnosis and treatment decisions.