The Clinical Value Of MRI Quantitative Analysis Of Liver Fat Content

ObjectiveUsing ~1H-MRS and T2~*corrected multi echo Dixon imaging techniques for quanti-fication of the liver fat content(HFC)in the nonalcoholic fatty liver disease(NAFLD),we analyse the relationship between the value of MRS-PDFF,MRI-PDFF and the histopathological results.To evaluate the value of ~1H-MRS and T2~*corrected multi echo Dixon technique in diagnosis of fatty liver disease and to analyze the feasible of MRI quantification analysis in patients with chronic hepatitis B(CHB)in the determination of liver fat content.Materials and methods:According to the fatty liver diagnostic criteria in the latest version of the 2010 Chinese《the Nonalcoholic Fatty Liver Disease Diagnosis and Treatment Guidelines》,11 cases of simple nonalcoholic fatty liver disease(NAFLD),excluding the heterogeneity ones and confirmed by biopsy,were enrolled in this study;Conforms to the diagnostic criteria of chronic hepatitis b viral hepatitis of Chinese《the Chronic Hepatitis B Prevention and Cure Guidelines》,13 cases of simple chronic hepatitis B patients,which are also confirmed by biopsy,and 12 cases of CHB combined with NAFLD confirmed by biopsy were enrolled.A total of 36 patients,including 22 males and11 females,from 10 to 68 years old,with mean age 41.81±11.696 years old.All subjects were took the in-phase and out-phase imaging using the 1.5 T MR(Siemens,Magnetom Avanto),to judge whether the subjects had hepatic steatosis.Then they were underwent single voxel ~1H-MRS and T2~*corrected multi echo Dixon imaging sequences,to observe and analyze the MRS spectra features and the MRI images,to calculate the MRS-PDFF and MRI-PDFF values.Ultimately,They were all adopted the ultrasound guided percutaneous liver biopsy and pathological examination.The percentage of steatotic hepatocyte was measured.The interval time between these two operations was less than one week.In this study,we used the histopathologic results as the reference standard.The Spearman correlation analysis was first used in the simple NAFLD group,analyzing the correlation between MRS-PDFF,MRI-PDFF and the pathological findings.If the ~1H-MRS and T2~*corrected multi echo Dixon imaging techniques are effectively valuable in this group,the correlation between MRS-PDFF,MRI-PDFF and the pathological findings in CHB combined with NAFLD group will be evaluated,then to compared them with the simple NAFLD group.On the contrary,if the above imaging technology has no value in the simple NAFLD group,the operation will be aborted.Of the above three groups of patients,we used the single factor analysis of variance(ANOVA)to analyse the age difference among them.Because gender is classification variables,the chi-square test is adopted to analyze the gender difference among these three groups of patients.As the pathological findings are nonparametric variables,we used nonparametric test of two independent samples test among the three groups,we respectively analysed the difference between each two of the three groups.The difference of MRS-PDFF and MRI-PDFF among three groups of patients was analysed by the two comparison of single factor analysis of variance.Meanwhile,simple box plot was used to analyse whether there was a significant difference of MRS-PDFF and MRI-PDFF among these three groups of patients.Through the comparative analysis above,if there was no statistically significant difference between the simple NAFLD and CHB combined with NAFLD,we combined these two groups as the experimental group,while the CHB group as normal control.We used the ROC curve respectively to analyse the MRS-PDFF and MRI-PDFF.When the sensitivity and specificity were all very high,we then estimated the threshold of MRS-PDFF and MRI-PDFF.Results1.The correlation coefficient of MRS-PDFF and magnetic resonance imaging(MRI)-PDFF with pathological in simple fatty liver disease patients was 0.851(P=0.001)and0.833(P=0.001),respectively.They were all highly positive correlation.In Hepatitis B with fatty liver disease group,the correlation coefficient of MRS-PDFF and MRI-PDFF with pathological was 0.594(P=0.042)and 0.594(P=0.002),respectively.It also showed positive correlation.2.The age(P=0.604>0.05)and gender(P=0.110>0.05)among the three groups of patients had no significant differences.3.Only simple hepatitis B was significant difference between the other two groups(simple nonalcoholic fatty liver disease and hepatitis b with fatty liver group)of patients with pathological results,MRS-PDFF and MRI-PDFF(P value were all less than 0.05).However,there had no statistically significant between simple fatty liver disease and CHB with NAFLD group in the pathological results(P=0.294>0.05),MRS-PDFF(P=0.643>0.05)and MRI-PDFF(P=0.038>0.05).Meanwhile,we used simple box plot to analyse the difference among three groups of patients.The same results as described above were obtained.4.Because there had no difference between simple NAFLD and CHB combined with NAFLD,We concluded that the area under the ROC curve of MRS-PDFF and MRI-PDFF was 0.980 and 0.973,respectively.The estimated threshold of MRS-PDFF was 4.64%(sensitivity 95.7%and specificity 92.3%)and MRI-PDFF was 5.1%(sensitivity was 91.3%and specificity was 92.3%).Conclusion:~1H-MRS and T2*corrected multi echo Dixon techniques can be used for accurate quantitative analysis of the HFC in patients with NAFLD.Further observation shows that 1H-MRS and T2*corrected multi echo Dixon technique can be used for quantitative analysis of HFC in patients with chronic hepatitis B,and as the effective method in quantification of fatty liver disease.