MR多参数成像诊断肝纤维化分期的价值

Experimental study on early diagnosis of liver fibrosis using multi-parametric MRI

目的探讨MR弹性成像(MRE)、磁敏感加权成像(SWI)和T1ρ成像诊断肝纤维化(LF)分期的价值。方法将80只新西兰白兔采用随机区组分组法分为LF组(n=60)和对照组(n=20)。LF组按0.1 ml/kg每周皮下注射50%四氯化碳油溶液,第4、5、6、15周末任意选取LF组12只及对照组5只行轴面肝脏扫描,获得肝脏弹性硬度值(LS)、肝脏/肌肉信号比(SIR肝/肌)和T1ρ值,取肝组织进行病理学Scheuer纤维化分期,并将LF组和对照组分为不同群组。采用单因素方差分析评价不同LF分期组间LS、T1ρ和SIR肝/肌值的差异,采用Spearman法进行各特征值与病理分期的相关性分析,采用ROC评价三种MR成像方法(单一方法,结合两种方法、结合3种方法)诊断LF分期的效能。结果55只兔纳入研究,F0期14只、F1期11只、F2期10只、F3期9只、F4期11只。不同LF分期组间,LS值、SIR肝/肌及T1ρ值差异均有统计学意义(P<0.05)。LS值与LF分期呈正相关(r=0.910,P<0.01);SIR肝/肌值与LF分期呈负相关(r=-0.808,P<0.01);T1ρ值与LF分期呈正相关(r=0.512,P<0.01)。MRE诊断效能最高,诊断≥F1、≥F2、≥F3、F4期的ROC下面积分别为0.953、0.949、0.986、0.979;结合两种成像方法,MRE+SWI诊断≥F1、≥F2、≥F3、F4期LF的ROC下面积分别为0.965、0.983、0.991、0.950;3种MR特征值结合在诊断≥F1、≥F2、≥F3、F4期的ROC下面积分别为0.969、0.985、0.996、1.000。结论MRE作为诊断效能最高的手段,与SWI组合在肝纤维化早期诊断优于其他两两组合,结合3种方法能发挥最佳的诊断效能。

Objective To assess diagnostic performance of multi-parametric MRI including MR elastography (MRE), susceptibility-weighted imaging (SWI) and T1ρ imaging in detecting the early stage of liver fibrosis (LF). Methods Eighty healthy rabbits were randomly divided into LF group (n=60) and control group (n=20). The LF group (n=12) and control group (n=5) were randomly selected at the end of the 4th, 5th, 6th, 15th week after injection of 50% CCl4 oil solution, respectively. All rabbits underwent 3.0 T MRI scans and histopathological Scheuer staging. Differences between groups were examined using one-way analysis of variance with Dunnett's T3 test. Spearman correlation was used to analyze the correlation between liver stiffness (LS), liver-to-muscle SI ratio (SIR), T1ρ value in different LF stages. ROC curve analysis was used to assess the diagnostic performance. Results Fifty-five rabbits were included in our study, which covered F0 (n=14), F1 (n=11), F2 (n=10), F3 (n=9) and F4 (n=11). Significant differences of all characteristic values were found among different LF stages (P<0.05).There were significant differences in LS values between F0 and F2, F3, F4, respectively;F1 and F3, F4, respectively;F2 and F4;F3 and F4 (P<0.05). There were significant differences in SIR between F0 and F2, F3, F4, respectively;F1 and F2, F3, F4, respectively;F2 and F4 (P<0.05). T1ρ value showed significant differences between F0, F1, F2, F3 and F4, respectively (P<0.05). LS, SIR, and T1ρ values were correlated with LF stage (r=0.910,-0.808, 0.512, respectively, P<0.01). The area under curve (AUC) for LS value were greater than those for the other two methods (0.953 for≥F1, 0.949 for≥F2, 0.986 for≥F3, 0.979 for F4). The AUCs of the combination of MRE and SWI for detecting≥F1 was 0.965, for≥F2 was 0.983, for≥F3 was 0.991, and for F4 was 0.950. Combining all three MR methods showed the highest diagnostic performance for staging LF with AUCs of 0.969, 0.985, 0.996 and 1.000. Conclusion MRE is the most prominent MRI method, and combination of MRE and SWI show higher diagnostic performance than the others for staging LF;however, combining all three MR methods exhibits the most excellent diagnostic efficacy.