Liver kinetic studies of [18F]2-fluoro-2-deoxy-D-glucose (FDG) metabolism are important tools for functional information modeling and hepatic glucose metabolic rate estimates. Kinetic parameter estimates require the arterial blood time-activity curve and the tissue time-activity curve as input and output functions. For liver studies, however, the conventional arterial-input may not be consistent with the real model input because the liver has a dual blood supply from the hepatic artery and the portal vein to the liver. This study compares model parameter estimates using a dual-input function with results using an arterial-input function. The results show that the two methods provide different parameter estimates with the dual-input function leading to more accurate parameter estimates.