Application of Advanced Monte Carlo Methods in Numerical Dosimetry

Many tasks in different sectors of dosimetry are very complex and highly sensitive to changes of the radiation field, e. g. dose estimations in microdosimetry. Often experimental measurements are only available for particular conditions or at single target points. Thus only the simulation of radiation transport is capable of describing the radiation field completely. Different approaches for solving the radiation transport problem as e.g. FEM can be used. In particular the Monte Carlo (MC) method is very useful for high-dimensional problems, and it is accepted as a standard method for this purpose.
Down to sub cellular dimensions the energy deposition by cascades of secondary electrons is the main pathway for damage induction in matter. A high number of interactions is taking place until such an electron is slowed down to thermal energies. For this reason the non-analog Monte Carlo program AMOS has been developed for photon and electron transport. The advanced MC algorithms implemented are able to handle a large number of histories at reasonable performance even for time consuming single scattering models. They facilitate the calculation of photon dose distributions at up to 10^6 points.
Modeling the dose distribution of a 125-I brachytherapy source has proven the high efficiency of this MC approach. Further on results achieved by AMOS for the transport of electrons with low energies (E < 100 keV) in matter with low atomic numbers were compared to measured data from the literature. It has been proven that the program code produces results in rather well agreement to the experiments also for small target structures of about one micrometre. This is especially important for the application to cell irradiation experiments as they are carried out at the FZ Rossendorf. As another example of application a simulation of the whole spectral detector response of an HPGe detector will be presented.