Dosimetric evaluation and Monte Carlo simulation of a new proposed surface brachytherapy mould

The aim of this study was to develop a new in-house low-cost surface mould, and to evaluate its performance and dosimetric properties for high-dose-rate (HDR) cobalt-60 (⁶⁰Co) brachytherapy.

A water-equivalent surface mould was developed using medical silicone. Mould performance and dosimetry characteristics were evaluated with Monte Carlo N-particle (MCNP2.6) simulation, Gafchromic™ EBT3 film measurements, and treatment planning system (TPS) output. Three sample moulds with different thicknesses (i.e., 0.5 cm, 1 cm, and 1.5 cm) were constructed, and a phantom study was performed. Treatment plans prescribing 3 Gy to 0.5 cm under pseudo-skin were designed, and film dosimetry was completed. TPS dose distributions were compared using Monte Carlo (MC) simulation and film dosimetry.

Good consistency was observed between TPS results and film dosimetry at the prescribed depth of 0.5 cm, with mean differences of 0.70%, 0.40%, and 0.19% for mould thicknesses of 0.5 cm, 1 cm, and 1.5 cm, respectively. However, higher discrepancies were found at the phantom surface with 1.00%, 0.80%, and 0.56% dose differences for the considered mould thicknesses, respectively. These increased differences could be due to a higher dose gradient at the phantom surface, and a greater impact of uncertainties on the obtained results in this part. Moreover, mean differences between the results obtained from MC simulations and output of TPS at the prescribed depth of 0.5 cm were 0.73%, 0.60%, and 0.08% for mold thicknesses of 0.5 cm, 1 cm, and 1.5 cm, respectively. Higher variations were observed between TPS and MC at the phantom surface with 1.30%, 0.70%, and 0.13% dose differences for the considered mould thicknesses, respectively.

The developed surface mould demonstrated water equivalence at ⁶⁰Co energies, and was consistent with TPS calculations at routine treatment depths. Its effectiveness in non-melanoma skin cancer (NMSC) lesion treatment is highlighted. However, due to mould attenuation, TG-43-based TPS overestimated the dose delivered using this mould, especially at pseudo-skin surface, emphasizing the necessity for a model-based TPS algorithm.