A dedicated tool for PET scanner simulations using FLUKA

Positron emission tomography (PET) is a well-established medical imaging technique. It is based on the detection of pairs of annihilation gamma rays from a beta + -emitting radionuclide, usually inoculated in the body via a biologically active molecule. Apart from its wide-spread use for clinical diagnosis, new applications are proposed. This includes notably the usage of PET for treatment monitoring of radiation therapy with protons and ions. PET is currently the only available technique for non-invasive monitoring of ion beam dose delivery, which was tested in several clinical pilot studies. For hadron-therapy, the distribution of positron emitters, produced by the ion beam, can be analyzed to verify the correct treatment delivery.The adaptation of previous PET scanners to new environments and the necessity of more precise diagnostics by better image quality triggered the development of new PET scanner designs. The use of Monte Carlo (MC) codes is essential in the early stages of the scanner design to simulate the transport of particles and nuclear interactions from therapeutic ion beams or radioisotopes and to predict radiation fields in tissues and radiation emerging from the patient. In particular, range verification using PET is based on the comparison of detected and simulated activity distributions. The accuracy of the MC code for the relevant physics processes is obviously essential for such applications. In this work we present new developments of the physics models with importance for PET monitoring and integrated tools for PET scanner simulations for FLUKA, a fully-integrated MC particle-transport code, which is widely used for an extended range of applications (accelerator shielding, detector and target design, calorimetry, activation, dosimetry, medical physics, radiobiology, . . . ). The developed tools include a PET scanner geometry builder and a dedicated scoring routine for coincident event determination. The geometry builder allows the efficient construction of PET scanners with nearly arbitrary parameters. The scoring output can be saved in standard output formats, including list mode and binary sinogram, which facilitates the processing of the data with external reconstruction algorithms. We also present recent developments of Flair, the GUI for FLUKA, which allow to read DICOM files and convert them into FLUKA voxel geometry in a convenient way.