z-logo
Premium
Performance evaluation of a collapsed cone dose calculation algorithm for HDR Ir‐192 of APBI treatments
Author(s) -
Terribilini Dario,
Vitzthum Veronika,
Volken Werner,
Frei Daniel,
Loessl Kristina,
Veelen Bob,
Manser Peter,
Fix Michael K.
Publication year - 2017
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1002/mp.12490
Subject(s) - brachytherapy , imaging phantom , monte carlo method , dosimetry , nuclear medicine , formalism (music) , physics , biomedical engineering , mathematics , radiation therapy , medicine , radiology , statistics , art , musical , visual arts
Purpose Most dose calculations for HDR brachytherapy treatments are based on the AAPM ‐ TG 43 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 ‐ TG 43 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 ‐ TG 43 algorithm. Dose distributions calculated with ACE 4.5, ACE 4.5/corr, AAPM ‐ TG 43 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 ‐ TG 43 calculated dose distributions in a large water phantom are < 1%. When using HU s scaled densities for the breast cancer patients, both accuracy levels of ACE 4.5 overestimate the MC calculated dose distributions for all analyzed dosimetric parameters. In the planning target volume ( PTV ), ACE 4.5 ( ACE 4.5/corr) overestimates on average V 100%, PTV by 3% ± 1% (1% ± 1%) and D 50, PTV by 3% ± 1% (1% ± 1%) and in the organs at risk D 1cc, skin by 4% ± 2% (1% ± 1%), D 0.5cc, ribs by 4% ± 2% (0% ± 1%), and D 1cc, heart by 8% ± 2% (3% ± 1%) compared to MC . Comparisons of the individual dose components reveals an agreement for the primary component of < 2% local differences for both ACE 4.5 and ACE 4.5/corr. Local differences of about 40% (20%) for the first and residual scatter components where observed when using ACE 4.5 ( ACE 4.5/corr). Using uniform densities for one case shows a better agreement between ACE 4.5 and MC for all dosimetric parameters considered in this work. Conclusions In general, on the 10 APBI patients the ACE 4.5/corr algorithm results in similar dose distributions as the commonly used AAPM ‐ TG 43 within the PTV . However, the accuracy of the ACE 4.5/corr calculated dose distribution is closer to MC than to AAPM ‐ TG 43. The differences between commercial version ACE 4.5 and MC dose distributions are mainly located in the first and residual scatter components. In ACE 4.5/corr, the changes done in the algorithm for the scatter components substantially reduce these differences.

This content is not available in your region!

Continue researching here.

Having issues? You can contact us here