摘要
目的:探讨基于增强CT对肝细胞肝癌微血管侵犯(MVI)及其分级的术前预测价值。方法:回顾性分析167例经手术病理证实为肝细胞肝癌(HCC)患者的数据资料,按7︰3将数据分为训练组(118例)和验证组(49例)。验证组MVI阳性68例,其中低危(M1)组41例和高危(M2)组27例;MVI阴性(M0)50例。通过分析得出影响HCC患者MVI的重要危险因素,建立临床-影像模型并验证其诊断效能。结果:MVI临床-影像特征包括天冬氨酸转氨酶、血小板与淋巴细胞比值、肿瘤最大径、肿瘤边缘、瘤周强化、瘤内动脉、瘤周低密度。临床-影像特征模型对M0、M1、M2的诊断效能ROC曲线下面积(AUC)值分别为0.87、0.68、0.88,测试组相应的AUC值分别为0.81、0.59、0.79。结论:基于增强CT建立的临床-影像模型在术前无创地评估和预测MVI的发生及分级具有一定的临床应用价值。
Purpose:To evaluate the value of enhanced CT in preoperative prediction of microvascular invasion(MVI)and its grading in hepatocellular carcinoma(HCC).Methods:A total of 167 patients with HCC confirmed by surgical pathology were retrospectively analyzed,and the data were divided into training group(118 cases)and verification group(49 cases)according to 7︰3 ratio.In the validation group,68 cases were MVI positive,including 41 cases in the low-risk(M1)group and 27 cases in the high-risk(M2)group,and 50 cases were MVI negative(M0).The important risk factors for MVI in HCC patients were analyzed,and the clinical-imaging model was established and its diagnostic efficacy was verified.Results:The clinical-imaging features of MVI included aspartate aminotransferase,platelet and lymphocyte ratio,tumor maximum diameter,tumor margin,peritumoral enhancement,intratumoral artery,and peritumoral hypodensity.The area under ROC curve(AUC)values of clinical-imaging model for M0,M1 and M2 were 0.87,0.68 and 0.88,respectively,and the AUC values in test group were 0.81,0.59 and 0.79,respectively.Conclusion:The clinical-imaging model based on enhanced CT has certain clinical application value in noninvasive assessment and prediction of the occurrence and grading of MVI.
作者
郑鑫
徐运军
黄京城
陈迪
葛亚琼
ZHENG Xin;XU Yunjun;HUANG Jingcheng;CHEN Di;GE Yaqiong(Department of Radiology,Anhui Provincial Hospital Affliated to Anhui Medical University;Department of Radiology,The First fliated Hospital of Dalian Medical University;GE Healthcare China(Shanghai))
出处
《中国医学计算机成像杂志》
CSCD
北大核心
2022年第6期596-602,共7页
Chinese Computed Medical Imaging
基金
中央高校基本科研业务费专项资金资助(WK9110000062)。
关键词
计算机体层成像
肝细胞肝癌
微血管侵犯
Computed tomography
Hepatocellular carcinoma
Microvascular invasion
