体素不相干运动DWI预测肝癌微血管侵犯的价值

Value of Intravoxel Incoherent Motion Diffusion Weighted Imaging in Predicting Microvascular Invasion of Hepatocellular Carcinoma

目的探讨扩散加权成像（DWI）单指数模型和体素内不相干运动（IVIM）双指数模型对预测肝细胞癌（HCC）微血管侵犯的价值。方法搜集31例HCC患者的影像和临床资料,共34个病灶按病理分为微血管侵犯阳性组（18个）和微血管侵犯阴性组（16个）,利用单因素及多因素分析对比两组临床数据、表观扩散系数（standard ADC）值、灌注分数（F）值、真实扩散系数（slow ADC）值和灌注相关扩散系数（fast ADC）值,通过受试者工作特征（ROC）曲线评价参数诊断效能。结果微血管侵犯阳性组standard ADC、slow ADC、F、fast ADC分别为（1.15±0.26）×10^-3mm^2/s、（0.76±0.16）×10^-3mm^2/s、（29.43±11.61）%、（40.95±20.39）×10^-3mm^2/s;微血管侵犯阴性组standard ADC、slow ADC、F、fast ADC分别为（1.44±0.38）×10^-3mm^2/s、（0.97±0.23）×10^-3mm^2/s、（30.77±15.05）%、（45.49±25.36）×10^-3mm^2/s。微血管侵犯阳性组standard ADC和slow ADC低于微血管侵犯阴性组,且组间存在统计学差异（P=0.014,0.005）,而fast ADC和F值亦减低,但组间无统计学差异。单因素分析两组在肿瘤直径及分化、standard ADC和slow ADC有统计学差异,多因素回归分析仅slow ADC独立影响因素（P=0.03）,当slow ADC值取阈值1.02×10^-3mm^2/s,敏感度和特异度分别为94.4%、56.2%。结论 IVIM-DWI在术前预测HCC微血管侵犯有一定帮助,且slow ADC诊断效能最好。

Objective to explore the value of Mono-exponential and IVIM model of diffusion weighted imaging for predicting microvascular invasion（ MVI） of hepatocellular carcinoma（ HCC）. Methods The clinical and imaging characteristics of 31 patients with 34 HCC nodules confirmed by pathology were retrospectively collected. All patients underwent multi-b-fator DWI scan. The patients were divided into MVI group（ n = 16,16 nodules） and without MVI group（ n = 15,18 nodules）. Standard ADC、slow ADC、fast ADC、fraction of fast ADC 、clinical date were compared between MVI group and without MVI group based on univariate and multiple logistic regression analysis. Receiver operating characteristic（ ROC） curves were generated to evaluate the diagnostic performance of each parameter. Results standard ADC values of MVI and without MVI was（ 1. 15 ± 0. 26） ×10^-3 mm^2/s and（ 1. 44 ± 0. 38） ×10^-3 mm^2/s,respectively. slow ADC values of MVI and without MVI was（ 0. 76 ± 0. 16） ×10-3 mm^2/s and（ 0. 97 ± 0. 23） ×10^-3 mm^2/s,respectively. These two parameters were significantly lower in MVI groups than without MVI groups（ P = 0. 014,0. 005）,respectively. Fast ADC values of MVI and without MVI was（ 40. 95 ±20. 39） ×10^-3 mm^2/s and（ 45. 49 ± 25. 36） ×10^-3 mm^2/s,respectively. F values of MVI and without MVI was（ 29. 43 ±11. 61） % and（ 30. 77 ± 15. 05） %,respectively. F values and Fast ADC values of MVI groups were lower than that without MVI groups,but the differences in F and fast ADC values were not statistically significant between the two groups. A univariate analysis showed tumor grade and diameters,standard ADC and slow ADC were significantly associated with MVI. A multiple logistic regression analysis showed that a lower slow ADC was independent predictors of MVI. Taking 1. 02 ×10^-3 mm^2/s cutoff values of slow ADC,the diagnostic sensitivity and specificity were 94. 4%,56. 2%. Conclusion Slow ADC based on IVIM model is the most valuable parameter and is able to predict microvascular invasion of hepatocellular carcinoma.