Performance evaluation of a collapsed cone dose calculation algorithm for HDR Ir-192 of APBI treatments

Purpose Most dose calculations for HDR brachytherapy treatments are based on the AAPM-TG43 formalism. Because patient's anatomy, heterogeneities and applicator shielding are not considered, the dose calculation based on this formalism is inaccurate in some cases. Alternatively, Collapsed Cone (CC) methods as well as Monte Carlo (MC) algorithms belong to the model based dose calculation algorithms, which are expected to improve the accuracy of calculated dose distributions. In this work, the performance of a CC algorithm, ACE in Oncentra Brachy 4.5 (ACE 4.5), has been investigated by comparing the calculated dose distributions to the AAPM-TG43 and MC calculations for 10 HDR brachytherapy accelerated partial breast irradiation treatments (APBI). Comparisons were also performed with a corrected version of ACE 4.5 (ACE 4.5/corr). Methods The brachytherapy source microSelectron mHDR-v2 (Elekta Brachytherapy) has been implemented in a MC environment and validated by comparing MC dose distributions simulated in a water phantom of 80 cm in diameter with dose distributions calculated with the AAPM-TG43 algorithm. Dose distributions calculated with ACE 4.5, ACE 4.5/corr, AAPM-TG43 formalism and MC for 10 APBI patients plans have then been computed and compared using HU scaled densities. In addition, individual dose components have been computed using ACE 4.5, ACE 4.5/corr and MC and compared individually. Results Local differences between MC and AAPM-TG43 calculated dose distributions in a large water phantom are