A radiomics nomogram for preoperative prediction of microvascular invasion risk in hepatitis B virus-related hepatocellular carcinoma
A radiomics nomogram for preoperative prediction of microvascular invasion risk in hepatitis B virus-related hepatocellular carcinoma
复制标题
乙型肝炎病毒相关肝细胞癌术前预测微血管侵犯风险的放射组学列线图
DOI：
10.5152/dir.2018.17467
复制
发表时间：
2018-05-01
影响因子：
2.1
通讯作者：
Liu, Li
中科院分区：
文献类型：
--
作者：
Peng, Jie;Zhang, Jing;Liu, Li
[方法]将304例符合条件的肝细胞癌患者随机分为训练组(184例)和独立验证组(120例)。收集门静脉期和动脉期肝细胞癌的CT数据，提取放射学特征。使用最小绝对收缩和选择算子算法，对训练集进行处理以降低数据维度、特征选择和放射组学签名的构建。然后，用多元Logistic回归分析建立了包括放射组学特征、放射学特征和甲胎蛋白(AFP)水平的预测模型，如放射组学诺模图所示。分析放射组学诺模图的辨别能力、刻度和临床实用性。使用放射组学正常图验证内部队列数据。结果放射组学特征与MVI状态显著相关(P<；0.001，两组均为队列)。个体化预测标准图中保留了包括放射组学特征、肿瘤边缘不光滑、低密度晕、内动脉和甲胎蛋白水平在内的预测因子。该模型在训练和验证队列中显示出良好的校正和判别能力(C指数[95%可信区间]：0.846[0.787-0.905]和0.844[0.774-0.915])。结论放射组学诺模图作为一种无创性的术前预测方法，对乙肝病毒相关性肝细胞癌患者的MVI状态具有良好的预测准确性。
PURPOSEWe aimed to develop and validate a radiomics nomogram for preoperative prediction of micro-vascular invasion (MVI) in hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC).METHODSA total of 304 eligible patients with HCC were randomly divided into training (n=184) and independent validation (n=120) cohorts. Portal venous arid arterial phase computed tomography data of the HCCs were collected to extract radiomic features. Using the least absolute shrinkage and selection operator algorithm, the training set was processed to reduce data dimensions, feature selection, and construction of a radiomics signature. Then, a prediction model including the radiomics signature, radiologic features, arid alpha-fetoprotein (AFP) level, as presented in a radiomics nomogram, was developed using multivariable logistic regression analysis. The radiomics nomogram was analyzed based on its discrimination ability, calibration, and clinical usefulness. Internal cohort data were validated using the radiomics nomogram.RESULTSThe radiomics signature was significantly associated with MVI status (P < 0.001, both cohorts). Predictors, including the radiomics signature, nonsmooth tumor margin, hypoattenuating halos, internal arteries, and alpha-fetoprotein level were reserved in the individualized prediction nomogram.The model exhibited good calibration and discrimination in the training and validation cohorts (C-index [95% confidence interval]: 0.846 [0.787-0.905] and 0.844 [0.774-0.915], respectively). Its clinical usefulness was confirmed using a decision curve analysis.CONCLUSIONThe radiomics nomogram, as a noninvasive preoperative prediction method, shows a favorable predictive accuracy for MVI status in patients with HBV-related HCC.