Liver Perfusion gives important information about blood supply of liver. However, in daily clinical diagnosis, radiologists have to manually mark the perfusion position in time-sequence images due to the motion of liver caused by respiration. In this paper, we propose a novel hybrid method using a variation of active contours and modified chamfer matching to automatically detect the liver perfusion position and measure the intensity with a single shape prior. The experiment is taken on abdomen MRI series and the result reveals that after extracting liver’s rough boundary by active contours, precise perfusion positions can be detected by the modified chamfer matching algorithm, and finally a refined intensity curve without respiration affection can be achieved.