基于超声影像组学的联合模型预测冷冻胚胎移植后妊娠结局的临床价值  

Clinical value of a combined model based on ultrasound radiomics for predicting pregnancy outcomes after frozen embryo transfer

作　　者：王琼仨 芦金飞 陈文艳 张丽丽 WANG Qiongsa;LU Jinfei;CHEN Wenyan;ZHANG Lili(Department of Ultrasound,Zhoushan Women and Children Hospital,Zhoushan 316100,China)

机构地区：[1]浙江省舟山市妇女儿童医院超声科,浙江舟山316100

出　　处：《临床超声医学杂志》2025年第4期332-339,共8页Journal of Clinical Ultrasound in Medicine

摘　　要：目的探讨基于超声影像组学的联合模型预测冷冻胚胎移植者妊娠结局的临床价值。方法选取于我院拟行冷冻胚胎移植的78例育龄期妇女,均于子宫内膜转化日行经阴道超声检查,获得子宫内膜回声分型、厚度、容积及相关血流参数。根据胚胎移植后妊娠结局分为妊娠阳性组(43例)和妊娠阴性组(35例),比较两组临床资料及超声参数的差异;采用多因素Logistic回归分析筛选预测冷冻胚胎移植者妊娠结局的影响因素,建立临床模型。将纳入对象以3∶1比例随机分为训练集(58例)和验证集(20例),应用PyRadiomics影像组学分析软件提取超声影像组学特征,采用最小冗余最大相关算法、最小绝对收缩和选择算子算法筛选及优化获得最优特征,计算超声影像组学评分(Rad-score)。基于上述影响因素和Rad-score构建联合模型并绘制列线图可视化。绘制受试者工作特征(ROC)曲线分析各模型预测冷冻胚胎移植者妊娠结局的效能;校准曲线分析模型的校准度;临床决策曲线分析模型的临床收益。结果妊娠阳性组与妊娠阴性组年龄、不孕类型、基础促卵泡生成素(FSH)、胚胎类型比较差异均有统计学意义(均P<0.05)。多因素Logistic回归分析显示,年龄、不孕类型、基础FSH、胚胎类型均为预测冷冻胚胎移植者妊娠结局的独立影响因素(均P<0.05),由此构建临床模型方程:Logit(P)=-0.234+0.500×年龄+0.700×不孕类型+0.600×基础FSH+1.200×胚胎类型。共提取479个超声影像组学特征,经过筛选及优化后最终获得9个最优特征,计算Rad-score作为超声影像组学模型。纳入年龄、胚胎类型、Rad-score构建联合模型。ROC曲线分析显示,临床模型、超声影像组学模型及联合模型预测训练集和验证集患者妊娠结局的曲线下面积(AUC)分别为0.66、0.70、0.82和0.60、0.72、0.81,以联合模型的AUC最高,与其他模型比较差异均有统计学意义(均P<0.05)�Objective To explore the clinical value of a combined model based on ultrasound radiomics for predicting pregnancy outcomes in patients undergoing frozen embryo transfer.Methods A total of 78 women of childbearing age who underwent frozen embryo transfer in our hospital were selected.All patients underwent transvaginal ultrasound examination on the day of endometrial transformation to obtain endometrial echo typing,thickness,volume,and related blood flow.Patients were divided into positive pregnancy group(43 cases)and negative pregnancy groups(35 cases)according to the pregnancy outcomes after embryo transfer.The differences in clinical data and ultrasound parameters between the two groups were compared.Multivariate Logistic regression analysis was used to identify influencing factors for predicting pregnancy outcomes in patients undergoing frozen embryo transfer,and a clinical model was established.Patients were randomly divided into a training set(58 cases)and a validation set(20 cases)at a ratio of 3∶1.Ultrasound radiomics features were extracted by the PyRadiomics radiomics analysis software.The minimum redundancy maximum relevance algorithm and the least absolute shrinkage and selection operator algorithm were used to select and optimize the best features,and the ultrasound radiomics score(Rad-score)was calculated.A combined model was constructed based on the above influencing factors and Rad-score,and visualized using a nomogram.Receiver operating characteristic(ROC)curve was drawn to analyze the predictive efficacy of each model for pregnancy outcomes in patients undergoing frozen embryo transfer.Calibration curve was used to analyze the calibration of the models,and decision curve was used to analyze the clinical benefits of the models.Results There were statistically significant differences in age,infertility type,basal follicle-stimulating hormone(FSH),and embryo type between the positive and negative pregnancy groups(all P<0.05).Multivariate Logistic regression analysis showed that age,infertility ty

关 键 词：超声检查 影像组学 列线图 冷冻胚胎移植 妊娠结局 

分 类 号：R445.1[医药卫生—影像医学与核医学]