Modeling of data from a dynamic acquisition is complicated by the combination of camera motion, organ motion, and the time-varying nature of the radionuclide distribution being imaged which results in inconsistent projection data sets. Our approach is to estimate the time varying radionuclide distribution directly from projection data. Entire tomographic slices or volumes are represented by a multiresolution spatiotemporal parameterization. The heart is modeled using spatiotemporal model that takes into account changes in the deformation of the heart as well as changes in the intensity of the signal caused by uptake and washout of the radiopharmaceutical from the heart and motion of the camera. We have developed an algorithm to reconstruct the continuous kinetics of the radiopharmaceutical by using spatiotemporal basis functions when the cardiac cycle of the heart is uniformly gated. The temporal basis functions are chosen to reflect the optimal tracer dynamics while the gate basis functions are assumed to be periodic with respect to the beating of the heart. Our preliminary results show that the fast gantry rotation of the SPECT camera can delineate the proper positioning of the myocardium in the reconstructed image while the ‘out of phase’ slow camera rotation smear out the motion of the heart.