Objective:To analyze the change rules of quantitative parameters of magnetic resonance-perfusion weighted imaging (MR-PWI) in cynomolgus monkeys with different degrees of liver fibrosis, and to explore the best parameter of MR-PWI in evaluating the severity of liver fibrosis.Methods:Liver fibrosis models of twenty-two cynomolgus monkeys were successfully established by subcutaneous injection of carbon tetrachloride and feeding with high-fat food. Among them, 15 cynomolgus monkeys developed into early liver cirrhosis (stage S4 of liver fibrosis). Compatibility group design was adopted, the comparative study on MR-PWI of exchange double blood supply model of liver was carried out in these 15 cynomolgus monkeys with a complete development process of liver fibrosis. The quantitative parameters of MR-PWI included endothelial transfer constant ( ktrans), reflux rate constant ( kep), extravascular extracellular space fractional volume ( ve), fractional plasma volume ( vp) and hepatic artery perfusion index (HPI). The change rules of the above parameters and their correlation with the severity of hepatic fibrosis were analyzed. The best parameter of MR-PWI was explored. Compatibility group design (randomized block design), analysis of variance, SNK- q test, Spearman rank correlation analysis and receiver operating characteristic (ROC) curve analysis were used for statistical analysis. Results:ktrans and kep of MR-PWI of cynomolgus monkeys decreased along with the progress of hepatic fibrosis, and the differences were statistically significant ( F=685.228, 99.718, both P<0.01). There were statistically significant differences between each stage of hepatic fibrosis (S1 to S4) and normal liver tissue (S0) ((0.527±0.038), (0.479±0.035), (0.432±0.032) and (0.387±0.031) mL/min vs.(0.584±0.044) mL/min, all P<0.01; (2.193±0.307), (1.997±0.301), (1.624±0.174) and (1.532±0.130) mL/min vs. (2.565±0.482) mL/min, all P<0.01). There were statistically significant in ktrans and kep between stage S3, S4 severe liver fibrosis and stage S1 mild liver fibrosis, stage S2 moderate liver fibrosis (all P<0.01), however there were no statistically significant differences between stage S3 and stage S4 liver fibrosis, between stage S1 and stage S2 liver fibrosis (all P>0.05). Along with the development of the severity of liver fibrosis, HPIs increased gradually, and the differences were statistically significant ( F=839.883, P<0.01). The HPIs of stage S0 to S4 were 0.244±0.022, 0.317±0.035, 0.421±0.046, 0.546±0.043 and 0.651±0.058, respectively, and there were statistically significant differences between groups (all P<0.01). Along with the progression of the severity of liver fibrosis, vp decreased while ve increased gradually, but there were no statistically significant differences among groups (all P>0.05). The results of Spearman rank correlation analysis indicated that ktrans and kep were negatively correlated with the severity of liver fibrosis ( rs=-0.875 and -0.797, both P<0.01), however HPI was positively correlated with the severity of liver fibrosis ( rs=0.959, P<0.01). The results of ROC curve analysis showed that area under curves (AUCs) of ktrans, kep and HPI in the diagnosis of early cirrhosis were 0.852 (95% CI 0.767 to 0.937), 0.799 (95% CI 0.700 to 0.897) and 0.967 (95% CI 0.932 to 1.002), respectively. The best cut-off values were 0.395 mL/min, 1.561 mL/min and 0.590, respectively. The sensitivity was 86.7%, 79.6% and 97.4%, respectively and the specificity was 77.4%, 71.9% and 93.1%, respectively. The thresholds of HPI in the diagnosis of liver fibrosis at stage S1, stage S2, stage S3 and stage S4 were 0.291, 0.376, 0.503 and 0.590, respectively; the sensitivity was 95.7%, 93.8% and 94.4% and 97.4%, respectively and the specificity was 89.5%, 84.7%, 91.3% and 92.7%, respectively. Conclusions:The parameters of MR-PWI change regularly with the development of liver fibrosis in the cynomolgus monkey model, among which HPI is the best parameter for quantitative evaluation of the severity of liver fibrosis.