双层探测器光谱CT在评估食管癌分化程度及脉管神经侵犯中的应用价值  

Application value of dual-layer spectral detector CT in evaluating differentiation and lymphovascular neural invasion of esophageal cancer

作　　者：郑欢欢 刘松[1] 朱妍 李茗[1] 张海龙 魏晓磊[1] ZHENG Huanhuan;LIU Song;ZHU Yan;LI Ming;ZHANG Hailong;WEI Xiaolei(Department of Radiology,Nanjing Drum Tower Hospital,the Affiliated Hospital of Nanjing University Medical School,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

出　　处：《东南大学学报（医学版）》2022年第4期510-516,共7页Journal of Southeast University(Medical Science Edition)

摘　　要：目的:探讨双层探测器光谱CT特征与手术切除的食管癌分化程度及脉管神经侵犯之间的关系,为治疗决策选择及预后评估提供依据。方法:回顾性分析47例经术后病理证实食管癌患者的术前光谱CT胸部增强图像并获得光谱CT特征参数,包括120 kVp混合能量图像、40~70 keV(间隔10 keV)的虚拟单能量图像(VMI)、碘密度、无水碘及有效原子序数图。不同分化程度间光谱CT特征的关系采用单因素方差分析。采用受试者工作特征(ROC)曲线分析光谱CT特征在鉴别食管癌不同分化程度(低vs.中-高)、有无脉管神经侵犯中的效能。结果:40~60 keV VMI、碘密度、无水碘浓度及有效原子序数值在食管癌低、中、高分化程度间差异具有统计学意义(P值分别为0.030、0.033、0.043、0.028、0.025及0.028)。光谱CT参数识别低分化程度及脉管神经侵犯的诊断效能较高,其中碘密度值识别肿瘤低分化的AUC最高(0.775),50 keV VMI鉴别有无神经侵犯的AUC最高(0.890),60 keV VMI区分有无脉管神经同时侵犯的AUC最高(0.804)。结论:双层探测器光谱CT特征可评估食管癌分化程度及脉管神经侵犯。碘密度对肿瘤低分化程度、50 keV VMI对神经侵犯、60 keV VMI对脉管神经同时侵犯的诊断价值最高。Objective:To explore the value of dual-layer spectral detector CT features of surgically resected esophageal cancer in evaluating differentiation and lymphovascular neural invasion,and to provide a basis for the selection of treatment decisions and the evaluation of prognosis.Methods:Preoperative chest spectral CT images were retrospectively evaluated in 47 surgically resected esophageal cancer and spectral CT features were obtained,including 120 kVp polyenergetic image,40~70 keV virtual monoenergetic image(VMI),iodine density,anhydrous iodine,and effective atomic number maps.One-way analysis of variance was used for assessing the relationship between spectral CT characteristics of different differentiation degrees.Receiver operating characteristic(ROC)curve was used to analyze the effectiveness of spectral CT characteristics in differentiating esophageal cancer with different differentiation levels(low vs.medium-high),lymphovascular and neural invasion.Results:40~60 keV VMI,iodine density,anhydrous iodine concentration,and effective atomic number values were significantly different among the low,medium,and highly differentiated esophageal cancers(P values were 0.030,0.033,0.043,0.028,0.025 and 0.028,respectively).Spectral CT parameters had high diagnostic efficiency in identifying poorly differentiated degrees and lymphovascular neural invasion.Among them,the iodine density had the highest AUC(0.775)for distinguishing poorly differentiated tumors,50 keV VMI had the highest AUC(0.890)for identifying neural invasion,and 60 keV VMI showed the highest AUC(0.804)for identifying concurrent lymphovascular nerve invasion.Conclusion:Spectral CT features of double-detector can be used to evaluate the degree of differentiation and lymphovascular neural invasion of esophageal cancer.The values of iodine density in the diagnosis of low differentiation,50 keV VMI in the diagnosis of neural invasion and 60 keV VMI in the diagnosis of concurrent lymphovascular neural invasion are the highest.

关 键 词：食管肿瘤 光谱CT 分化程度 脉管侵犯 神经侵犯 

分 类 号：R735.1[医药卫生—肿瘤] R814.42[医药卫生—临床医学]