DWI鉴别肝硬化背景下LI-RADS 4级肝癌与不典型增生结节

Differential diagnosis of LI-RADS-4 hepatocellular carcinoma and atypical dysplastic nodule under thebackground of liver cirrhosis using diffusion weighted imaging

目的:探讨定性及定量DWI对肝硬化背景下LI-RADS 4级肝癌与不典型增生结节的鉴别诊断价值。方法:回顾性分析经术前MRI检查诊断为肝脏影像报告及数据系统(LI-RADS)分类为4级且经病理证实的71例肝硬化结节患者的病例资料。共80个病灶,包括肝癌(HCC)52个和不典型增生结节(DN)28个。分析病灶的信号特征,测量病灶的ADC值,包括最小ADC值、最大ADC值、平均ADC值、ADC值标准差及相对ADC值(病灶与肝实质的平均ADC值的比值)。采用单因素分析比较各项参数在HCC组和DN组之间的差异,采用多因素logistic回归分析建立联合诊断模型,并采用ROC曲线分析评估模型的诊断效能。结果:HCC组与DN组之间DWI高信号、T_(2)WI高信号、最小ADC值、最大ADC值、平均ADC值及相对ADC值的差异有统计学意义(P<0.05),各变量的ROC曲线下面积(AUC)的取值范围为0.695~0.797。Logistic回归分析结果显示DWI高信号和相对ADC值是鉴别HCC与DN的独立影响因素,两者构建的联合诊断模型鉴别HCC与DN的AUC值为0.857(95%CI:0.770~0.945),敏感度为90.38%(47/52),特异度为78.57%(22/28),符合率为86.25%(69/80)。结论:DWI对肝硬化背景下LI-RADS 4级HCC与DN的鉴别诊断具有一定价值,多参数联合诊断可提高鉴别效能。

Objective:The purpose of this study was to investigate the value of qualitative and quantitative diffusion weighted imaging(DWI)in differentiating hepatocellular carcinoma(HCC)with Liver Imaging Report and Data System(LI-RADS,LR)-4 from dysplastic nodule(DN)in patients with liver cirrhosis.Methods:The clinical and imaging data of 71 patients with liver cirrhosis nodules categorized as LR-4 were retrospectively analyzed,based on the pathological results,a total of 80 lesions were included in the study(52 HCCs and 28 DNs).MRI signal characteristics of the lesions were analyzed,and ADC values of the lesions were measured on the apparent diffusion coefficient images(ADC),including the minimum,maximum and mean and standard deviation of ADC and the relative ADC(the ratio of the mean ADC value of the lesions to the liver parenchyma).The value of each parameter in differential diagnosis of HCC and DNs was analyzed by univariate analysis.The combined diagnostic model was constructed using multivariate logistic analysis.The diagnostic efficiency of the model was evaluated by ROC curve analysis.Results:There were statistically significant differences in high signal on DWI,high signal on T_(2)WI,minimum ADC value,maximum ADC value,average ADC value and relative ADC value between HCC group and DN group(P<0.05)by using univariate analysis,and the area under the ROC curve(AUC)ranged from 0.695 to 0.797.Multivariate analysis showed that high signal on DWI and relative ADC value were independent influencing factors for the differentiation of HCCs and DNs.When these two factors were combined to construct a diagnostic model for differentiating HCCs and DNs,The AUC was 0.857(95%CI:0.770~0.945),with the sensitivity of 90.38%(47/52),the specificity of 78.57%(22/28),and the accuracy of 86.25%(69/80).Conclusion:Qualitatively and quantitatively DWI has certain value in the differential diagnosis of LR-4 HCCs and DNs under the background of liver cirrhosis,and the combined model can improve the diagnostic efficiency.