The first thorny problem in the study of atrial fibrillation (AF) is the establishment of reliable animal models. This paper presents how to create a canine sterile pericarditis model. A new data analysis method, Fast Fourier Transform is introduced. The data from canine sterile pericarditis model was analyzed using FFT. The results demonstrate that canine sterile pericarditis model is a reliable animal model for studying AF and that FFT is a rapid and effective method for analyzing AF data.