DCE-CT联合直方图分析技术鉴别早期肝脓肿与肝内胆管细胞癌的价值分析  

The Value of DCE-CT Combined with Histogram Analysis Technique in Differentiating Early Liver Abscess from Intrahepatic Cholangiocarcinoma

作　　者：杨孟忱 郭昊翔 郭艺 郝跃文 刘海洋[1] YANG Meng-chen;GUO Hao-xiang;GUO Yi;HAO Yue-wen;LIU Hai-yang(Department of Imaging,Shangluo Central Hospital,Shangluo 726000,Shaanxi Province,China;Department of Imaging,Xi'an Children's Hospital,Xi'an 710043,Shaanxi Province,China)

机构地区：[1]商洛市中心医院医学影像科,陕西商洛726000 [2]西安市儿童医院影像科,陕西西安710043

出　　处：《中国CT和MRI杂志》2024年第9期107-110,共4页Chinese Journal of CT and MRI

基　　金：中国红十字基金会2020年度ICON科研基金项目(XM_HR_ICON_2020_10_8)。

摘　　要：目的探讨CT增强扫描联合直方图分析技术在鉴别早期肝脓肿与肝内胆管细胞癌(intrahepatic cholangiocarcinoma,ICC)的应用价值并验证其诊断效能。方法收集我院行CT增强扫描检查的77例早期肝脓肿及肝内胆管细胞癌患者的影像资料,再使用MaZda软件勾画病变典型图例作为ROI(region of interest)进行直方图数据分析,将差异显著的指标纳入logistic回归分析,评估其单一参数模型及联合模型,应用ROC(receiver operating characteristic)曲线评价各参数价值,计算AUC(area underthecurve)以验证其诊断效能。结果早期肝脓肿组与ICC组比较,临床资料及CT特征中的病变部位、有无向心性强化/肝包膜凹陷征/动脉期灌注异常、病变最大径、动静脉期强化差值和直方图参数中的skewness、kurtosis有统计学意义(P均<0.05);Logistic回归分析得出病变部位、有无向心性强化/动脉期灌注异常、动静脉期强化差值、skewness、kurtosis为独立影响因素。ROC曲线显示联合模型的鉴别效能最高,AUC为0.982,敏感度和特异度分别为0.905、0.971,高于任意单一参数模型。结论CT增强扫描联合直方图分析技术在鉴别早期肝脓肿与肝内胆管细胞癌上效能很高,有助于提高术前诊断准确率。Objective To investigate the value of CT enhancement scanning combined with histogram analysis in differentiating early liver abscess and intrahepatic cholangiocarcinoma(ICC)and to validate its diagnostic efficacy.Methods The imaging data of 77 patients suffering from early-stage liver abscess and intrahepatic cholangiocarcinoma who underwent CT enhancement scanning examination in our hospital were collected,and then the histogram data were analyzed by checking the maximum level of the lesion as the region of interest(ROI)using MaZda software.The imaging and histogram parameters with significant differences between groups were included in the binary Logistic regression analysis to evaluate the single-parameter model and the joint model.The value of each parameter was evaluated by using the receiver operating characteristic(ROC)curve,and the area under the curve(AUC)was calculated and validated for diagnostic efficacy.Results In comparison between the early liver abscess group and the ICC group,there was a statistically significant difference between the clinical data and CT features in terms of lesion location,presence of centripetal enhancement/hepatic pericardial notch sign/arterial phase perfusion abnormality,the maximum diameter of the lesion,the difference in enhancement between the arterial and venous phases,and Skewness,and Kurtosis in the histogram parameter(all P<0.05).Through binary Logistic regression,the analysis yielded lesion site,presence or absence of centripetal enhancement/arterial phase perfusion abnormality,arteriovenous phase enhancement difference,Skewness,and Kurtosis as independent influencing factors.The ROC curve revealed that the joint model had the highest discriminatory efficacy,with an AUC of 0.982 and sensitivity and specificity of 0.905 and 0.971,respectively,which was higher than that of the single-parameter model.Conclusion CT-enhanced scanning combined with histogram analysis is more effective in differentiating early liver abscess and intrahepatic cholangiocarcinoma,which contr

关 键 词：肝脓肿 肝内胆管细胞癌 增强扫描 计算机体层成像 直方图 

分 类 号：R575.4[医药卫生—消化系统] R735.7[医药卫生—内科学] R445.3[医药卫生—临床医学]