Non-contrast T1 dispersion versus Gd-EOB-DTPA-enhanced T1mapping for the risk stratification of non-alcoholic fatty liver disease in rabbit models

Purpose: To investigate the diagnostic efficacy of T1 rho dispersion and Gd-EOB-DTPAenhanced T1mapping in the identification of early liver fibrosis (LF) and non-alcoholic steatohepatitis (NASH) in a non-alcoholic fatty liver disease (NAFLD) rabbit model induced by a high-fat diet using histopathological findings as the standard reference. Methods: A total of sixty rabbits were randomly allocated into the standard control group (n = 12) and the NAFLD model groups (8 rabbits per group) corresponding to different high-fat high cholesterol diet feeding weeks. All rabbits underwent noncontrast transverse T1 rho mapping with varying spin-locking frequencies (FSL = 0 Hz and 500 Hz), native T1 mapping, and Gd-EOB-DTPA-enhanced T1 mapping during the hepatobiliary phase. The histopathological findings were assessed based on the NASH CRN Scoring System. Statistical analyses were conducted using the intraclass correlation coefficient, analysis of variance, multiple linear regression, and receiver operating characteristics. Results: Except for native T1, T1 rho, T1 rho dispersion, HBP T1, and oT1 values significantly differed among different liver fibrosis groups (F = 14.414, 18.736, 10.15, and 9.799, respectively; all P < 0.05). T1 rho, T1 rho dispersion, HBP T1, and oT1 values also exhibited significant differences among different NASH groups (F = 4.138, 4.594, 21.868, and 22.678, respectively; all P < 0.05). In the multiple regression analysis, liver fibrosis was the only factor that independently influenced T1 rho dispersion (R2 = 0.746, P = 0.000). Among all metrics, T1 rho dispersion demonstrated the best area under curve (AUC) for identifying early LF (>= F1 stage) and significant LF (>= F2 stage) (AUC, 0.849 and 0.916, respectively). The performance of oT1 and HBP T1 (AUC, 0.948 and 0.936, respectively) were better than that of T1 rho and T1 rho dispersion (AUC, 0.762 and 0.769, respectively) for diagnosing NASH. Conclusion: T1p dispersion may be suitable for detecting liver fibrosis in the complex background of NAFLD, while Gd-EOB-DTPA enhanced T1 mapping is superior to nonenhanced T1p mapping (T1p and T1p dispersion) for identifying NASH.