ROSI - an object-oriented and parallel-computing Monte Carlo simulation for X-ray imaging

In the field of X-ray imaging, Monte Carlo simulation is an important tool. It gives the possibility of understanding experimental results and it allows the construction of virtual imaging setups with predictions of their quality. For these reasons, we developed the Roentgen Simulation (ROSI) which is based on the object-oriented C++ class library GISMO. The interaction algorithms are based on the established EGS4-code and its current LSCAT-extension. ROSI introduces random variables for modelling physical parameters by a given random distribution, e.g. the source position or the direction and energy of the photons to be emitted. It is possible to run ROSI in parallel on a local computer network (Beowulf cluster) to obtain simulation data in shorter time. Finally, it has an easy-to-use interface. We will present the concept of ROSI and demonstrate its flexibility by an example. 2003 Elsevier Science. B.V. All rights reserved.