SU‐E‐T‐584: Evaluation of Collapsed Cone Dose Calculation Algorithm for HDR Brachytherapy Using Monte Carlo

Abstract

Purpose: In clinical routine, dose calculation for HDR brachytherapy plans is usually based on the AAPM‐TG43 formalism. Because the patient's anatomy is not taken into account, the resulting dose calculation may be inaccurate. Monte Carlo (MC) methods are able to calculate dose distributions accurately, but due to long computation times they are mainly used for benchmarking purposes. In this work, a test version of the collapsed cone (CC) algorithm from Nucletron was evaluated using MC. Methods: The Nucletron brachytherapy source micro Selectron‐HDR v2 has been implemented in the MC environment. CC and MC calculated dose distributions for a single source placed in a water phantom are validated against published AAPM‐TG43 data. Furthermore, dose distributions for clinical breast cancer patient treatment plans are calculated with CC and MC. Primary and scatter dose components are separated and compared individually. Results: The local differences between the MC calculated dose distributions in the water phantom and the AAPM‐TG43 data are less than 1%, while the differences between MC and CC for distances > 2 cm from the source are up to 2% of the reference dose (dose at 1 cm from the source in its equatorial plane). These differences are assigned to the first and residual scatter dose components. For the clinical breast cancer patients, CC underestimates the MC dose up to 2% in the high‐dose region (> 100% of the prescribed dose). This deviation is correlated with the scan‐length of the patient CT data set. If the CT data set extends the PTV by an appropriate amount, the underestimation falls below 1%. Conclusion: The CC algorithm is an accurate dose calculation algorithm and improves the dose calculation accuracy in brachytherapy compared to AAPM‐TG43 based algorithms. This work was supported by Nucletron. This work was supported by Nucletron.