Activation of Collimators Irradiated With Clinical Proton Beams and Development of a Semiempirical Model for Activity Calculation.

Patient-specific collimators used in proton therapy are activated after use. The aim of this work is to assess the residual activity in brass collimators considering clinical beams, so far studied only for monoenergetic beams, and to develop a model to calculate the activity. Eight brass collimators irradiated with different clinical and monoenergetic beams were included in the study. The collimators were analyzed with gamma spectrometry in the framework of three independent studies carried out at the two French proton therapy sites. Using FLUKA (a fully integrated particle physics Monte Carlo simulation package), simulations were performed to determine radionuclides and activities for all the collimators. The semiempirical model was built using data calculated with FLUKA for a range of clinical beams (different maximum proton energies, modulations, and doses). It was found that there was global coherence in experimental results from different studies. The relevant radionuclides at 1 mo postirradiation were Co, Co, and Zn, and additionally, Mn, Co, and Co for high-energy beams. For nondegraded monoenergetic beams, differences between FLUKA and spectrometry were within those reported in reference benchmark studies (±30%). Due to the use of perfect monochromatic sources in the FLUKA model, FLUKA results systematically underestimated experimental activities for clinical beams, especially for Zn, depending on the beam energy spread (modulation, degradation, beam line characteristics). To account for the energy spread, correction factors were derived for the semiempirical model. The model is applicable to the most relevant radionuclides and total amounts. Secondary neutrons have a negligible contribution to the activity during treatment with respect to proton activation.