基于多期CT影像组学鉴别卵巢良、恶性肿瘤的临床应用价值

Clinical application value of multi-phase CT radiomics in differentiating benign and malignant ovarian tumors

目的:探讨基于多期CT影像组学模型及联合模型鉴别卵巢良、恶性肿瘤的临床应用价值。方法:回顾性搜集2018年1月-2022年7月经术后病理证实的144例良性及182例恶性卵巢肿瘤患者的临床及CT多期影像组学资料,将患者随机按7:3分为训练组228例及验证组98例。图像预处理并利用ITK-SNAP勾画肿瘤病灶区域,用PY提取组学特征,将提取特征正则化,采用Spearman相关分析,相关系数大于0.9,保留其一特征,使用最小绝对收缩与选择算子(LASSO)回归分析对组学特征降维,建立影像组学标签。将影像组学的优势期相与临床指标结合建立联合模型,利用列线图分析预测效能。结果:筛选出平扫、动脉期、静脉期及延迟期的特征数分别为22、7、10、22个,构建各期影像组学模型,结果显示延迟期为最优模型,在训练组中,其受试者工作特征(ROC)曲线的曲线下面积(AUC)为0.857。将临床特征与延迟期影像组学特征构建联合模型,其ROC的AUC值为0.870,优于临床模型,且差异有统计学意义(Z=3.376,P=0.0007)。验证组中,联合模型ROC的AUC值为0.844,优于临床模型,差异无统计学意义(Z=1.650,P=0.0989)。结论:CT扫描的延迟期影像组学特征鉴别卵巢良性与恶性肿瘤的效能优于其它期相,延迟期影像组学标签与临床特征相结合的联合模型有较高的鉴别诊断效能。

Objective:To investigate the clinical application value of multi-phase CT radiomics model and combined model in differentiating benign and malignant ovarian tumors.Methods:The preoperative clinical data and multi-phase CT radiomics characteristics of 144 patients with benign ovarian tumors and 182 patients with malignant ovarian tumors confirmed by postoperative pathology from January 2018 to July 2022 were retrospectively collected.The patients were randomly divided into the training group(228 cases)and the validation group(98 cases)according to the ratio of 7:3.After image preprocessing,ITK-SNAP was used to delineate the tumor lesion area,and PY was used to extract the radiomics features.The extracted features were regularized.Spearman correlation analysis was used to calculate whether the correlation coefficient was greater than 0.9.Least absolute shrinkage and selection operator(LASSO)regression analysis was used to reduce the dimensionality of radiomics features and establish radiomics signature.The dominant phase of radiomics and clinical data were combined to establish a clinical combined model,and the nomogram was used to show the prediction efficiency.Results:The features of plain scan,arterial phase,venous phase and delay phase were screened out to be 22,7,10,22.The imaging radiomics models of each phase were constructed,and it was found the results showed that the delay phase was the optimal model.In the training group,the area under receiver operating curve(AUC)was 0.857.The combined model of clinical features and delayed stage imaging features was constructed,and the AUC was 0.870,which was better than the clinical model,and the difference was statistically significant(Z=3.376,P=0.0007).In the verification group,the AUC of the combined model was 0.844,better than that of the clinical model,and the difference was not statistically significant(Z=1.650,P=0.0989).Conclusion:The delayed phase radiomics features of CT scan are superior in differentiating benign and malignant ovarian tumors from other phases.The combined model of delayed phase radiomics features and clinical data has high value in clinical differential diagnosis.