Monte Carlo study of ionization chamber magnetic field correction factors as a function of angle and beam quality

Purpose: To use EGSnrc Monte Carlo simulations for magnetic field dosimetry to determine optimal measurement orientations, calculate beam quality conversion factors for 32 cylindrical and three parallel-plate (PP) ion chambers, evaluate the beam quality and angular dependence of these factors, and examine the magnetic field effects on % dd(10)(x) and TPR1020. Methods: Beam quality conversion factors, k(Q)(mag), and magnetic field conversion factors, k(B) = k(Q)(mag) /k(Q), are calculated as a function of chamber rotation for six cylindrical ionization chamber in either a Co-60 beam with a 0.35 T magnetic field or a 7 MV beam with a 1.5 T field, both magnetic fields are perpendicular to the photon beam. The chambers' sensitive air volumes are varied by either using the entire geometric volume or excluding the air volume associated with the first 1 mm away from the stem. The k(B) and k(Q)(mag) factors are evaluated using four clinical photon spectra. The variation in % dd(10)(x) and TPR1020 as a function of magnetic field for six photon spectra are studied using DOSXYZnrc. Results: When the magnetic field is perpendicular to the photon beam, orienting the chamber parallel with the magnetic field reduces the magnetic field effect on chamber response (i.e., dose to air per water dose) and variations due to the unknown sensitive volume are essentially eliminated. Calculated k(B) factors are within 1% of unity for the majority of cylindrical chambers, although larger k(B) values are associated with chambers with high-Z electrodes. PP chambers have k(B) corrections as large as 8.9% and have a larger angular sensitivity compared to cylindrical chambers. Values of k(B) for cylindrical ion chambers are independent of beam quality, except for chambers with high-Z electrodes. For % dd(10)(x) values between 63.3% and 73.8%, k(B) varies by at most (0.26 +/- 0.15)% when the magnetic field is perpendicular to the photon beam and parallel to the chamber. Differences in % dd(10)(x), between no magnetic field and with a 1.5 T field perpendicular to the photon beam are (0.04 +/- 0.10)%, (1.89 +/- 0.10)%, and (6.20 +/- 0.10)% for a Co-60, 7, and 25 MV photon beam, respectively, while TPR1020 shows less than (0.36 +/- 0.10)% change. Applying the ICRU-90 recommendations for stopping powers instead of ICRU-37 is found to change k(Q) (and hence k(B)) by less than 0.1%. Conclusions: Orienting the chamber parallel to the magnetic field when the field is perpendicular to the photon beam will minimize the effect of the magnetic field on chamber response, and eliminate the problem of the unknown sensitive volume. Values of k(B) and k(Q)(mag) can bring ion chamber dosimetry in magnetic fields in-line with the TG-51 protocol. PP chamber are sensitive to the magnetic field and variation in chamber response due to small angular changes makes them unlikely candidates for clinical reference dosimetry in magnetic fields. The stability in TPR1020, as a function of magnetic fields and beam qualities, makes it the best beam quality specifier in magnetic fields. (C) 2017 American Association of Physicists in Medicine