On the use of the absorbed depth-dose measurements in the beam calibration of a surface electronic high-dose-rate brachytherapy unit, a Monte Carlo-based study.

Purpose: To evaluate the use of the absorbed depth-dose as a surrogate of the half-value layer in the calibration of a high-dose-rate electronic brachytherapy (eBT) equipment. The effect of the manufacturing tolerances and the absorbed depth-dose measurement uncertainties in the calibration process are also addressed.

Methods: The eBT system Esteya (Elekta Brachytherapy, Veenendaal, The Netherlands) has been chosen as a proof-of-concept to illustrate the feasibility of the proposed method, using its 10 mm diameter applicator. Two calibration protocols recommended by the AAPM (TG-61) and the IAEA (TRS-398) for low-energy photon beams were evaluated. The required Monte Carlo (MC) simulations were carried out using PENELOPE2014. Several MC simulations were performed modifying the flattening filter thickness and the x-ray tube potential, generating one absorbed depth-dose curve and a complete set of parameters required in the beam calibration (i.e., HVL, backscatter factor (B ), and mass energy-absorption coefficient ratios (µ /ρ) ), for each configuration. Fits between each parameter and some absorbed dose-ratios calculated from the absorbed depth-dose curves were established. The effect of the manufacturing tolerances and the absorbed dose-ratio uncertainties over the calibration process were evaluated by propagating their values over the fitting function, comparing the overall calibration uncertainties against reference values. We proposed four scenarios of uncertainty (from 0% to 10%) in the dose-ratio determination to evaluate its effect in the calibration process.

Results: The manufacturing tolerance of the flattening filter (±0.035 mm) produces a change of 1.4% in the calculated HVL and a negligible effect over the B , (µ /ρ) , and the overall calibration uncertainty. A potential variation of 14% of the electron energies due to manufacturing tolerances in the x-ray tube (69.5 ± ~10 keV) generates a variation of 10% in the HVL. However, this change has a negligible effect over the B and (µ /ρ) , adding 0.1% to the overall calibration uncertainty. The fitting functions reproduce the data with an uncertainty (k = 2) below 1%, 0.5%, and 0.4% for the HVL, B , and (µ /ρ) , respectively. The four studied absorbed dose-ratio uncertainty scenarios add, in the worst-case scenario, 0.2% to the overall uncertainty of the calibration process.

Conclusions: This work shows the feasibility of using the absorbed depth-dose curve in the calibration of an eBT system with minimal loss of precision.