An Experiment-Based Approach for Predicting Positron Emitter Distributions Produced During Therapeutic Ion Irradiation

In-beam positron emission tomography (PET) is a valuable method for a beam-delivery independent dose monitoring in radiation therapy with ion beams. Up to now, its clinical feasibility has been demonstrated for patient irradiation with carbon ions. From radiobiological point of view it is highly desirable to perform tumor irradiation also with other light ions. To extend the application of in-beam PET also to these ions, extensive knowledge about positron emitter production via nuclear fragmentation reactions during ion irradiation is necessary. To model the positron emitter production correctly, cross sections for all possible nuclear reactions occurring in the tissue during irradiation and leading to positron emitters are required. Since these cross sections are available only for a few reaction channels in the required energy range, a novel approach for estimating the positron emitter production from experimental data is introduced. The prediction of positron emitter distributions is based on depth dependent thick target yields, which are derived by linear superposition of measured yields in water, graphite and polyethylene as reference materials. Results on the prediction of positron emitter distributions in polymethyl methacrylate (PMMA) as well as inhomogeneous targets induced by lithium and carbon irradiation are presented. By comparison with data deduced from experiments, it is shown that a rather accurate prediction of positron emitter distribution is feasible using this method.