Abstract: A method and imaging system for creating a motion compensated 3D volumetric derived image from a 4D volumetric medical image is described. The method comprising the steps of: acquiring a 4D volumetric medical image comprising a plurality of 3D volumetric medical images forming a time-series of 3D volumetric medical images representing phases of a physiological motion cycle; determining dominant motion of each 3D volumetric medical image with respect to a reference location that represents a position relative to the motion cycle; resampling each of the 3D volumetric medical images to the reference location to compensate for the determined dominant motion; estimating secondary motion in at least one of each motion compensated 3D volumetric medical images or of each of the original 3D volumetric medical images; and constructing a 3D volumetric derived image by combining data from the motion compensated 3D volumetric images based on the estimated secondary motion.