Investigating the accuracy of the FLUKA code for transport of therapeutic ion beams in matter.

In-beam positron emission tomography (PET) is currently used for monitoring the dose delivery at the heavy ion therapy facility at GSI Darmstadt. The method is based on the fact that carbon ions produce positron emitting isotopes in fragmentation reactions with the atomic nuclei of the tissue. The relation between dose and beta(+)-activity is not straightforward. Hence it is not possible to infer the delivered dose directly from the PET distribution. To overcome this problem and enable therapy monitoring, beta(+)-distributions are simulated on the basis of the treatment plan and compared with the measured ones. Following the positive clinical impact, it is planned to apply the method at future ion therapy facilities, where beams from protons up to oxygen nuclei will be available. A simulation code capable of handling all these ions and predicting the irradiation-induced beta(+)-activity distributions is desirable. An established and general purpose radiation transport code is preferred. FLUKA is a candidate for such a code. For application to in-beam PET therapy monitoring, the code has to model with high accuracy both the electromagnetic and nuclear interactions responsible for dose deposition and beta(+)-activity production, respectively. In this work, the electromagnetic interaction in FLUKA was adjusted to reproduce the same particle range as from the experimentally validated treatment planning software TRiP, used at GSI. Furthermore, projectile fragmentation spectra in water targets have been studied in comparison to available experimental data. Finally, cross sections for the production of the most abundant fragments have been calculated and compared to values found in the literature.