Objective: To investigate the value of magnetic resonance imaging (MRI) in predicting the type and severity of placenta accreta (PA). Methods: The clinical data of 67 PA cases in Peking University Third Hospital from January, 2015 to November, 2016 were analyzed retrospectively.Cases were classified according to the direct and indirect MRI signs.Chi-square test was applied to disclose the difference between different PA types and MRI signs.MRI signs with statistical significance were further calculated by the area under ROC curve (AUC) value analysis to reveal the correlation between specific signs and the type/severity of PA. Results: Among the 67 cases, 24 cases were of mild PA and 43 cases were of severe PA.Statistical difference was found between the two PA groups in some signs such as placenta tissues invading the uterus myometrium, placenta tissues extending beyond the uterus outline, the uniformity of placental signal, abnormal vessel imaging in the placenta basilar part, Bladder Tent Sign and the discontinuity of low signal zone in the gap between bladder and placenta.AUC analysis showed that the above MRI signs had the stronger correlations between the type/severity of PA, compared with other signs. Conclusion: Not only can MRI be used to predict the type of PA, but also it can be used to predict the location and severity of PA, which has an important reference value for clinical works.
目的： 探讨核磁共振(MRI)预测胎盘植入类型及严重程度的价值。 方法： 回顾性分析2015年1月至2016年11月在北京大学第三医院分娩的67例胎盘植入患者的临床资料；根据MRI影像学直接和间接征象对各病例进行归类，比较各指标间的差异，以有统计学意义的MRI征象计算其ROC曲线下面积(AUC)，比较各种征象在判断胎盘植入类型中的相关性。 结果： 67例中轻型24例，重型43例，两组MRI影像学指标在胎盘组织侵入子宫肌层、胎盘组织突出于子宫轮廓外、胎盘信号不均匀、胎盘基底部血管影、膀胱帐篷征、膀胱胎盘间隙低信号带不连续指标上差异有统计学意义。AUC分析显示，上述指标与植入的严重程度相关性高于其他指标。 结论： MRI不仅可以在产前预测胎盘植入类型，而且可以判断植入部位及程度，对临床诊治有重要参考价值。.
Keywords: Forecasting; Magnetic resonance imaging; Placenta accreta; Prenatal diagnosis; Severity of illness index.