T1rho relaxation has been demonstrated applicable on the diagnosis of liver fibrosis in rat models [1] and also on the diagnosis of human liver cirrhosis by previous research studies[2,3]. But one recent study[4] found T1rho was a potential biomarker of liver function instead of liver fibrosis. The purpose of this study is to find out the relationship among T1rho, liver fibrosis stages and liver function in CCl4 induced liver fibrosis using rat models. Liver fibrosis of different stages were induced by intraperitoneal injection of 40% carbon tetrachloride (CCl4) with the dose of 2ml/kg, twice a week in 41 male Sprague Dawley rats. Another 12 rats were injected the same dose of saline. 5 to 6 rats of fibrosis group and 2 rats of control group were scanned 1,2,4,6,8,10 weeks after the injection started. MRI scans including T1rho imaging was performed on a PHILIPS 3.0T clinical scanner (Achieva TX, Best, The Netherlands), with an animal coil. T1rho was performed with a 3D fast field echo (FFE) sequence. Scan parameters were as follows: spin lock time 0,10,20,40,60ms, spinlock frequency 500Hz, TR/TE 4.9ms/2.4ms, FOV 80mm×80mm×10mm, flip angle 400, number of slices 5, NSA 4 respectively. T1rho images were fitted using IDL (Research Systems, Inc., USA) on a pixel-by-pixel basis to the exponentially decaying T1rho function to generate the T1rho relaxation map. Blood was drawn from right atrium before execution for a serum laboratory test. Liver fibrosis of rats were staged according to the stage system for rodents (F0~6), using pathological sections stained with Masson’s trichrome, picrosirius red and hematoxylin and eosin.All rats in the control group survived and were staged as F0. In the fibrosis group, 9 rats died, 32 rats were scanned and their fibrosis stages were: F0=2, F1=5, F2=5, F3=5, F4=5, F5=5, F6=5. T1rho values of the liver for F0~F6 were 40.68±1.60ms, 42.80±1.08ms, 44.98±1.04ms, 46.47±3.47ms, 47.73±3.84ms, 55.38±7.46ms and 57.08±5.11ms, respectively(Fig 1). Statistical difference was found in the T1rho value among different fibrosis stage groups (P<0.001). One-way ANOVA showed statistical differences between F0 and F2~6, F1 and F3~6, F2 and F5~6, F3 and F5~6, F4 and F5~6; no statistical differences were found between F0 and F1, F1 and F2, F2 and F3~4, F3 and F4, F5 and F6. Correlation was significant between T1rho values and liver fibrosis stage and correlation coefficient was 0.883 (P<0.0001). Of all blood serum parameters, AST, ALT, ALP were found positively correlated with both liver fibrosis stages and T1rho values; ALB was found negatively correlated with both liver fibrosis stages and T1rho values. No significant correlation was found between GLU, TBIL, T-CHO and T1rho values of the liver and liver fibrosis stages (table 1).Prior research studies supposed T1rho was just a potential biomarker for liver function, instead of a diagnosis method for liver fibrosis [4]. Our research confirmed the value of T1rho in the diagnosis of liver fibrosis in rat models. It could be used to diagnose F2-6 liver fibrosis. We also uncovered that correlation was significant both between T1rho values of liver and blood serum parameters and between liver fibrosis stages and blood serum parameters, and the correlation coefficient between liver fibrosis stages and blood serum parameters was usually higher. This may illustrate that the pathology changes of liver fibrosis or liver cirrhosis lead to abnormal liver function, which results in the changes in blood serum parameters. But as blood serum parameters are indirect biomarkers of liver fibrosis and could be affected by many other factors, it’s unstable compared with T1rho. We could also get images of whole liver by using T1rho scan instead of only numbers. T1rho could be use to diagnose liver fibrosis, and it’s also a potential biomarker of liver function.