Application of a Yield Approach for the Prediction of Positron Emitter Distributions Produced During Therapeutic Carbon-Ion Beam Irradiation

Positron emission tomography (PET) is a clinically implemented method for in vivo verification of treatment delivery in ion-beam therapy. The dose distribution is the relevant measure but PET captures a beta(+)-activity distribution, that is created as a side product during the irradiation. Therefore, a prediction of the activity distribution is required. We present a simulation code that makes use of measured yields of beta(+)-emitters to perform this task. The yields are available in three reference materials. A conversion from X-ray gray values to a combination of these materials has been set up. The lateral properties of the ion beam are described by means of a formulation of Moliere's scattering theory in thick targets. A model of the annihilation point density around an emitting nuclide has been implemented in the software to describe the movement of the positrons. Simulations of phantom experiments as well as of a real patient case were performed. The results are comparable or superior to the results obtained from an established condensed history Monte Carlo code. We conclude that measured yields can be used as raw data for the calculation of ion-beam-induced beta(+)-activity distribution instead of nuclear cross sections.