联合多期增强CT纹理分析及血液学指标术前预测胃癌VEGFR2表达状态  

Preoperative prediction of VEGFR2 expression in gastric cancer:integrating texture analysisbased on multi-phase contrast-enhanced CT and hematological parameters

作　　者：王玉婷 谭玲玉 朱小倩 刘祥[1] 李琳[2] 刘松[1] WANG Yuting;TAN Lingyu;ZHU Xiaoqian;LIU Xiang;LI Lin;LIU Song(Department of Radiology,Nanjing Drum Tower Hospital,Affiliated Hospital of Medical School,Nanjing University,Nanjing 210008,China;Department of Pathology,Nanjing Drum Tower Hospital,Affiliated Hospital of Medical School,Nanjing University,Nanjing 210008,China.)

机构地区：[1]南京大学医学院附属鼓楼医院医学影像科,江苏南京210008 [2]南京大学医学院附属鼓楼医院病理科,江苏南京210008

出　　处：《医学影像学杂志》2024年第3期57-61,共5页Journal of Medical Imaging

摘　　要：目的探讨多期动态增强CT纹理分析参数联合血液学指标术前预测胃癌血管内皮生长因子受体2(vascu‐lar endothelial growth factor receptor2,VEGFR2)表达状态的价值。方法选取本院148例胃癌患者资料,获得术前血液学指标和三期增强CT纹理分析参数。基于组内相关系数和差异性检验对参数进行特征筛选。基于二元Logistic回归构建血液学模型、CT纹理分析模型及综合模型来预测VEGFR2表达状态。通过受试者工作特征曲线评估三个模型的诊断效能,并通过列线图来可视化地预测胃癌患者VEGFR2的表达状态。结果基于血液学指标构建的血液学模型曲线下面积(area under the curve,AUC)为0.687。由静脉期纹理分析参数构建的CT纹理分析模型的AUC值为0.624。联合血液学模型和CT纹理分析模型构建的综合模型AUC值为0.723。结论联合多期动态增强CT纹理分析参数及血液学指标有助于术前预测胃癌VEGFR2表达状态。Objective To investigate the value of multi-phase dynamic contrast-enhanced CT texture analysis parameters combined with hematological parameters in preoperative prediction of vascular endothelial growth factor receptor 2(VEGFR2)expression in gastric cancer(GC).Methods This retrospective study included 148 patients with GC in our hospital.The preop‐erative hematological parameters and three-phase contrast-enhanced CT texture analysis parameters were obtained.The param‐eters were screened based on the intraclass correlation coefficient and difference test.The hematologic model,CT texture analy‐sis model,and integrated model were built based on binary logistic regression to predict the expression status of VEGFR2.The di‐agnostic efficacy of the three models was evaluated usingthe receiver operating characteristic curves,and the expression status of VEGFR2 in gastric cancer patients was visually predicted by nomogram.Results The area under the curve(AUC)of the hema‐tological model builtby hematological parameters was 0.687.The AUC value of the CT texture analysis model based on the venous phase was 0.624.Combined with the hematology model and CT texture analysis model,the AUC value of the integrated model was 0.723.Conclusion The combination of multi-phase dynamic contrast-enhanced CT texture analysis parameters and hema‐tological parameters can help us to predict the expression status of VEGFR2 in gastric cancer preoperatively.

关 键 词：胃癌 血管内皮生长因子受体2 体层摄影术 X线计算机 纹理分析 

分 类 号：R735.2[医药卫生—肿瘤] R814.42[医药卫生—临床医学]