Preoperative evaluation of liver function is crucial in patients candidate to surgical management, especially in those with liver fibrosis¹. Current approaches for assessing liver function mainly include Child-Pugh classification, MELD score, indocyanine green (ICG) plasma clearance test, 3-dimensional CT volumetry, and 99mTc-GSA SPECT. However, most of them are unable to evaluate regional liver function, and the spatial resolution of 99mTc-GSA SPECT is limited, so none of these methods can fully meet the clinical requirements_^2,3. Gd-EOB-DTPA is a hepatocyte-specific magnetic resonance imaging (MRI) contrast agent, the uptake and metabolism of which is related to hepatocyte function⁴. The aim of this study was to investigate the potential of Gd-EOB-DTPA enhanced MRI [including dynamic contrast-enhanced (DCE) images, as well as multiple hepatocyte-phase and T1 mapping images] in evaluating hepatic function in rats with liver fibrosis, and to discover specific imaging biomarkers for hepatic function. Forty male Sprague-Dawley rats were used in this study. Liver fibrosis in rats was induced by carbon tetrachloride intraperitoneal injection for 4-12 weeks (n=30). In control group (n=10) normal saline was applied. At 4, 8 and 12 weeks, rats from both control group and experiment group were randomly selected to undergo magnetic resonance (MR) exams. MR exams were performed using a 3.0T scanner (Siemens MAGNETOM Skyra) with a wrist coil. The MR protocol contained DCE sequence, multiple hepatocyte-phase acquisitions and T1 mapping series. TWIST-VIBE sequence was used for DCE imaging, in which sixty measurements including 3 pre-contrast ones were taken at a time interval of 3 s up to 3 min after the injection of the contrast agent. For multi-hepatocyte-phase images, single phase TWIST-VIBE sequence was used. Twelve datasets were acquired at 5, 10, 15, 20, 25, 30, 35, 45, 50, 55 and 60min after the injection of Gd-EOB-DTPA. Double Flip Angel T1 3D VIBE sequence was used for T1 mapping imaging. Seven series were acquired before and at 10, 20, 30, 40, 50 and 60min after the injection. Data were analyzed using a commercial available post-processing workstation (Siemens MMWP, tissue 4D). The following parameters were obtained: 1) liver perfusion parameters such as maximum relative enhancement (RE_max) and the time of maximum RE (T_max); 2) hepatocyte-phase parameters such as relative enhancement (RE) at different time point and elimination half-life of RE (T_RE1/2); 3) T1 mapping parameters such as T1 reduction rate at different time point (△T1%) and elimination half-life of △T1% (T_△T1%1/2). The ICG clearance test was performed using 0.5 mg/kg ICG after MR exams. Plasma ICG retention rate at 15 minutes after intravenous injection of ICG (ICG R15) was calculated. METAVIR score was used for pathological grading of liver fibrosis: normal (F0), mild fibrosis (F1-2), and advanced fibrosis (F3-4). SPSS 21.0 was used for statistical analysis. Pearson correlation coefficient assessment was used to evaluate the relationship between the ICG R15 values and the MR parameters. Twenty-two rats completed ICG clearance test in the end of the study, including normal (n=8), mild fibrosis (n=9) and advanced fibrosis (n=5). Nineteen of these 22 rats completed DCE-MR and multiple hepatocyte-phase acquisitions, while 10 completed T1 mapping acquisitions. ICG R15 was significantly higher in advanced fibrosis than normal and mild fibrosis (P=0.011, P=0.006). T_max and T_RE1/2 increased as the ICG R15 increased, the correlation coefficient with ICG R15 were 0.531 (P=0.019) and 0.535 (P=0.018) respectively. T_△T1%1/2, △T1%_30min, △T1%_40min, △T1%_50min increased as the ICG R15 value increased, the correlation coefficient were 0.697 (P=0.02) , 0.811 (P=0.033), 0.673 (P=0.043), 0.757 (P=0.028) respectively (Fig 1, Fig 2). RE_max and RE at different time points did not show significant correlation with ICG R15. Liver function decreased as liver fibrosis progressed. The time of maximum RE, half elimination time of RE and half elimination time of T1 reduction rate increased as liver function decreased, indicating a slower uptake of Gd-EOB-DTPA and a longer excretion time in rats with impaired liver function. Besides, at hepatocyte-phase, rats with liver dysfunction tend to retent more contrast agent in liver parenchyma, resulting in a higher T1 reduction rate at 30、40、50min. Parameters calculated from Gd-EOB-DTPA enhanced MRI such as T_max , T_RE1/2, T_△T1%1/2, △T1%_30min, △T1%_40min, and △T1%_50min exhibited moderate to high correlation with ICG R15 in rats with liver fibrosis. Gd-EOB-DTPA enhanced MRI could be helpful for the evaluation of liver function.