The invention relates to a computer-implemented method of processing image frames of a sequence of cardiac images using a neural network system, the method comprising:sequentially receiving a first sequence of image frames at a first neural network of the neural network system, the first sequence of image frames comprising a sequence of images of a cardiovascular system；outputting from the first neural network a first set of feature values, the first set of feature values comprising a plurality of data subsets, each corresponding to a respective image frame of the first sequence of image frames and relating to spatial features of the respective image frame； receiving the first set of feature values at a second neural network of the neural network system, different from the first neural network；outputting from the second neural network a second set of feature values relating to temporal features of the spatial features； andperforming a determination process to determine, based on the second set of feature values, a cardiac phase value relating to a cardiac phase associated with a first image frame of the first sequence of image frames.
