Objective To investigate the feasibility of gadoxetic acid disodium(Gd-EOB-DTPA)enhanced MR image signal intensity and imaging histology for assessing liver function.Methods The data of 105 patients with liver cirrhosis were analyzed,and the patients were classified into Child-Pugh class A(62 patients),Child-Pugh class B(38 patients),and Child-Pugh class C(5 patients)according to the Child-Pugh scoring system.The relative degree of enhancement(RE)of the liver,the liver-spleen signal ratio(SIliver/spleen)in the hepatobiliary phase,the liver-vertical muscle signal ratio(SIliver/muscle)inthehepatobiliary phase and the bile duct signal intensity contrast ratio(SIR)of the patients were meas-ured,and the imaging histological features of the images in the hepatobiliary phase were extracted using MaZda software,and the images were analysed by Fisher,POE+ACC,and MI three methods to downscale the extracted feature parameters,and then Spearman correlation analysis to exclude redundant parameters.The diagnostic efficacy of each signal intensity and im-aging features was quantitatively evaluated by using the receiver operating characteristic curve(ROC)curve,and multivari-ate Logistic regression analysis was constructed based on these parameters.Results The areas under the curve of RE,Sliver/spleen,SIliver/muscle and SIR were 0.805,0.740,0.630,and 0.703,respectively,to discriminate Child-Pugh A from Child-Pugh B+C.The parameters of the imaging histological characteristics,Perc.10％,S(0,5)Correlat,and Teta4 were effective in discrimination between Child-Pugh A and Child-Pugh B+C groups.Child-Pugh A and Child-Pugh B+C groups were 0.852,0.807,and 0.754,respectively.The area under the curve of the constructed multivariate Logistic regression analysis model was 0.961,with 95％CIs ranging from 0.929 to 0.993.Conclusion Based on the Gd-EOB-DTPA-enhanced MR images of the various signal intensity parameters and imaging histological features,the liver function,and the model con-structed by the combination of the two can better assess liver function.
Gd-EOB-DTPASignal intensity parametersImaging histologyLiver function
NETL
NSTL
万方数据
