Development and assessment of new imaging technologies requires the use of images that contain realistic anatomical structures. For example in developing a computer-aided detection scheme a large number of patient images are required. This is often not possible for new technologies that are underdevelopment. Further, optimization of new x-ray imaging systems are best done by modeling the physics of the system. If done accurately, the model can predict the appearance of the final displayed image, given an input object. That is, the model can produce simulated images. This paper discusses different approaches to produce simulated x-ray images and how they relate to system optimization.