多参数MRI影像组学列线图模型可有效预测宫颈鳞癌淋巴结转移  

Multi-parameter MRI nomogram model can effectively predict lymph node metastasis of cervical squamous cell carcinoma

作　　者：孟影 王书芹 张志雅 刘信信 岳凤辉 傅文悦 朱广辉 MENG Ying;WANG Shuqin;ZHANG Zhiya;LIU Xinxin;YUE Fenghui;FU Wenyue;ZHU Guanghui(Department of Radiology,The First Affiliated Hospital of Bengbu Medical University,Bengbu 233004,China;Department of Radiology,Fengyang County People's Hospital,Chuzhou 233100,China)

机构地区：[1]蚌埠医科大学第一附属医院放射科,安徽蚌埠233004 [2]凤阳县人民医院放射科,安徽滁州233100

出　　处：《分子影像学杂志》2025年第2期197-204,共8页Journal of Molecular Imaging

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淋巴结转移的准确预测。方法回顾性收集2018年12月~2021年11月于蚌埠医学院第一附属医院就诊、病理确诊宫颈鳞癌的216例患者的临床、病理及影像资料。按照7∶3的比例将所有患者随机分为训练组(n=151)、验证组(n=65),在训练组患者选取矢状面T2WI、T1WI增强及横轴位扩散加权成像图像,在病灶最大层面边缘勾画获取感兴趣区提取影像特征,应用最小绝对收缩选择算子(LASSO)算法建立影像组学评分。采用多因素Logistic回归分析确定临床分期、鳞状细胞癌相关抗原、淋巴结短径为独立危险因素,结合影像组学评分建立MRI影像组学列线图模型,并用验证组数据对列线图模型进行验证。运用ROC曲线下面积(AUC)评价模型的预测性能。应用校正曲线及决策曲线评估列线图模型的临床应用价值。结果基于临床参数及影像组学评分构建的列线图模型(AUC=0.912)的诊断效能高于临床特征模型(AUC=0.872)及影像组学模型(AUC=0.777)。结论联合临床模型和影像组学评分构建的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淋巴结转移的方法。Objective To develop and validate an MRI radiomics nomogram model for the accurate preoperative prediction of lymph node metastasis in patients with cervical squamous cell carcinoma.Methods The clinical,pathological and imaging data from patients diagnosed with cervical squamous cell carcinoma at the First Affiliated Hospital of Bengbu Medical College from December 2018 to November 2021 were retrospectively collected.All patients were randomly allocated into a training cohort(n=151)and a validation cohort(n=65)following a 7:3 ratio.Within the training cohort,sagittal T2WI,T1WI and transverse diffusion-weighted images were selected.Regions of interest were delineated at the margins of the largest lesions to extract imaging features.The LASSO algorithm was employed to construct the radiomics score.Multivariate logistic regression analysis identified FIGO stage,squamous cell carcinoma associated antigen levels,and lymph node short-axis diameter as independent risk factors.An MRI radiomics nomogram model was developed by integrating radiomics scores with clinical parameters,and this nomogram model was validated using an independent validation cohort.The area under the curve(AUC)was utilized to assess the predictive performance of the model.Calibration curves and decision curves were employed to evaluate the clinical utility of the nomogram model.Results The nomogram model incorporating clinical parameters and radiomics scores(AUC=0.912)demonstrated superior diagnostic efficacy compared to the clinical feature model(AUC=0.872)and the radiomics model alone(AUC=0.777).Conclusion The integration of MRI radiomics with clinical parameters into a nomogram model represents a simple,effective,and reliable approach for predicting lymph node metastasis in cervical squamous cell carcinoma.

关 键 词：宫颈鳞癌 MRI 列线图 淋巴结转移 

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