基于不同时相增强CT的影像组学对胰腺实性假乳头状肿瘤侵袭性行为的预测价值  

Multiphasic enhanced CT-based radiomics signature for preoperatively predicting the invasive behavior of pancreatic solid pseudopapillary neoplasm

作　　者：黄文鹏 刘思耘 李莉明[1] 韩懿静 梁盼[1] 吕培杰[1] 高剑波[1] Huang Wenpeng;Liu Siyun;Li Liming;Han Yijing;Liang Pan;Lyu Peijie;Gao Jianbo(Department of Radiology,the First Affiliated Hospital of Zhengzhou University,Zhengzhou 450052,China;Precision Health Institute(China),GE Healthcare,Beijing 100176,China)

机构地区：[1]郑州大学第一附属医院放射科,郑州450052 [2]GE医疗精准医学研究院(中国),北京100176

出　　处：《中华放射学杂志》2022年第1期55-61,共7页Chinese Journal of Radiology

基　　金：国家自然科学基金(81971615)。

摘　　要：目的:探讨CT影像组学对胰腺实性假乳头状肿瘤(pSPN)侵袭性行为的预测价值。方法:回顾性分析2012年1月至2021年1月郑州大学第一附属医院经术后病理证实的pSPN患者的CT图像,其中侵袭性23例、非侵袭性59例。分别在平扫、动脉期和静脉期CT图像上逐层勾画感兴趣区(ROI)获得三维ROI,每个ROI提取1 316个组学特征。将数据集经随机分层抽样法按照7∶3的比例分为训练集和验证集,在训练集中采用200%样本合成过采样技术(SMOTE)算法进行过采样,生成侵袭性和非侵袭性平衡数据用于建立训练模型,将构建的模型在验证集中进行验证。通过受试者操作特征(ROC)曲线分析评估不同模型的预测性能,并通过Delong检验比较不同模型的曲线下面积(AUC)值,采用连续净重分类改善度(NRI)和综合区分改善度(IDI)评估不同模型对分类效能的改善能力。结果:经过特征筛选,保留2、6、3个特征分别构建平扫、动脉期和静脉期模型,基于单独期相与联合时相模型共建立7个模型,除平扫模型外,其他模型预测pSPN侵袭性的AUC均>0.800。单期相中动脉期模型具有最优的鉴别效能,在SMOTE训练集和验证集的AUC值分别为0.913和0.873。在联合期相模型中,动脉-静脉期模型的AUC在训练集和验证集中最高,为0.934和0.913。训练集和验证集中,平扫-动脉-静脉期联合模型的AUC与动脉-静脉期模型的AUC差异无统计学意义( P均>0.05)。在验证集中,相较动脉-静脉期联合模型,进一步联合平扫不能获得正向分类改善(NRI、IDI均<0)。 结论:动脉期CT影像组学模型具有较好的术前预测pSPN侵袭性的性能,联合动脉期和静脉期的影像组学模型可进一步提高模型的性能。Objective To explore the value of multiphasic CT-based radiomics signature in predicting the invasive behavior of pancreatic solid pseudopapillary neoplasm(pSPN).Methods The multiphasic CT images of patients with pSPN confirmed by postoperative pathology in the First Affiliated Hospital of Zhengzhou University from January 2012 to January 2021 were analyzed retrospectively.There were 23 cases of invasiveness and 59 cases of non-invasiveness.The region of interest(ROI)was artificially delineated layer by layer in the plain scan,arterial-phase and venous-phase images,respectively.The 1316 image features were extracted from each ROI.The data set was divided into training and validation sets with a ratio of 7∶3 by stratified random sampling,and synthetic minority oversampling technique(SMOTE)algorithm was used for oversampling in the training set to generate invasive and non-invasive balanced data for building the training model.The constructed model was validated in the validation set.The receiver operating characteristic(ROC)analysis was used to evaluate model performance and the Delong′s test was applied to compare the area under the ROC curve(AUC)of different predict models.The improvement for classification efficiency of each independent model or their combinations were also assessed by net reclassification improvement(NRI)and integrated discrimination improvement(IDI)indices.Results After feature extraction,2,6 and 3 features were retained to construct plain-scanned model,arterial-phase and venous-phase models,respectively.Seven independent-phase and combined-phase models were established.Except the plain-scanned model,the AUC values of other models were greater than 0.800.The arterial-phase model had the best efficiency for classification among all independent-phase models.The AUC values of arterial-phase model in the SMOTE training and validation sets were 0.913 and 0.873,respectively.By combining the radiomics signature of the arterial-phase and venous-phase models,the AUC values of training and validati

关 键 词：胰腺肿瘤 体层摄影术 X线计算机 影像组学 实性假乳头状瘤 

分 类 号：R735.9[医药卫生—肿瘤] R730.44[医药卫生—临床医学]