基于MRI影像组学列线图在宫颈鳞癌组织学分级中的预测研究  

Prediction of Histological Grading of Cervical Squamous Cell Carcinoma Based on MRI Histograms

作　　者：孟影 刘信信 张志雅 岳凤辉 傅文悦 朱广辉[1] MENG Ying;LIU Xinxin;ZHANG Zhiya(Department of Radiology,The First Affiliated Hospital of Bengbu Medical College,Bengbu,Anhui Province 233004,P.R.China)

机构地区：[1]蚌埠医学院第一附属医院放射科,233004 [2]蚌埠医学院第二附属医院放射科,233030

出　　处：《临床放射学杂志》2023年第4期677-682,共6页Journal of Clinical Radiology

摘　　要：目的建立并验证MRI影像组学列线图模型,实现术前对宫颈鳞癌组织学分级的准确预测。方法回顾性搜集2019年1月至2021年10月于蚌埠医学院第一附属医院就诊208例患者的临床及影像资料。按照7∶3的比例将所有患者随机分为训练组(n=145)、验证组(n=63),在训练组患者选取矢状位T_(2)WI、增强T_(1)WI及轴位DWI图像,在病灶最大层面边缘勾画获取感兴趣区(ROI)提取影像特征,应用最小绝对收缩和选择算子(LASSO)算法建立影像组学评分。采用多因素Logistic回归分析确定独立危险因素,并结合影像组学评分建立MRI影像组学列线图。运用受试者工作特征曲线(ROC)曲线下面积(AUC)评价模型的预测性能。应用校正曲线评估列线图的临床应用价值。结果基于临床参数及影像组学评分构建的列线图模型(AUC:0.852)的诊断效能高于临床特征模型(AUC:0.723)及影像组学模型(AUC:0.788)。结论结合临床模型和影像组学评分的MRI影像组学列线图模型是一种简单、有效、可靠的预测宫颈鳞癌组织学分级的方法。Objective To establish and verify MRI image histograph model to accurately predict the histological grade of cervical squamous cell carcinoma before surgery.Methods Clinical and imaging data of patients admitted to the First Affiliated Hospital of Bengbu Medical College from January 2019 to October 2021 were retrospectively collected.All patients were randomly divided into the training group(n=145)and the validation group(n=63)in a 7∶3 ratio.Sagittal T_(2) WI,T_(1) WI enhancement(T_(1) CE)and transverse DWI images were selected for the training group,and the region of interest(ROI)image features were delineated at the edge of the largest plane of the lesion.The minimum absolute contraction se⁃lection operator(LASSO)algorithm was used to establish the image omics score.Multivariate Logistic regression analysis was used to determine the independent risk factors,and graph of MRI image omics were established in combination with the image omics score.The area under receiver operating Characteristic curve(AUC)was used to evaluate the prediction per⁃formance of the model.Calibration curve was used to evaluate the clinical value of rosette.Results The diagnostic efficacy of the rosette model(AUC:0.852)based on clinical parameters and imaging score was higher than that of the clinical fea⁃ture model(AUC:0.723)and the imaging model(AUC:0.788).Conclusion MRI radiomic line map model combined with clinical model and radiomic score is a simple,effective and reliable method to predict the histological grade of cervical squamous cell carcinoma.

关 键 词：宫颈鳞癌 磁共振成像 影像组学 列线图 组织学分级 

分 类 号：R737.33[医药卫生—肿瘤] R445.2[医药卫生—临床医学]