Improved heterogeneity handling for the Advanced Collapse-cone Algorithm
PD-0495
Abstract
Improved heterogeneity handling for the Advanced Collapse-cone Algorithm
Authors: Freja Alpsten1, Bob van Veelen2, Christian Valdes-Cortez3, Francisco Berumen Murillo4, Anders Ahnesjö5, Åsa Carlsson Tedgren6
1Karolinska University Hospital , Medical Radiation Physics and Nuclear Medicine, Stockholm, Sweden; 2Elekta Brachytherapy, Physics and Advanced Development, Veenendaal, The Netherlands; 3Hospital Regional de Antofagasta, Nuclear Medicine, Antofagasta, Chile; 4Université Laval, Département de physique, de génie physique et d'optique, Quebec, Canada; 5Uppsala University, Department of Immunology, Genetics and Pathology, Uppsala, Sweden; 6Karolinska University Hospital, Medical Radiation Physics and Nuclear Medicine, Stockholm, Sweden
Show Affiliations
Hide Affiliations
Purpose or Objective
The Advanced Collapsed-cone Engine (ACE) is a brachytherapy model-based dose calculation algorithm (MBDCA) developed for dose computations in heterogeneous media. However, it has been shown that ACE underestimates the dose to cortical bone, likely caused by shortcomings in the handling of the multiple scatter dose. The aim of this work was to investigate dose depositions with respect to scatter order in different media, and to test a correction to ACE in which the multiple-scatter (msc) dose component has been modeled in more detail with respect to the balance between local energy absorption and further transport.
Material and Methods
We calculated dose distributions with the current version of ACE, an updated version of ACE (ACEcorr), and with the TG43 formalism. To benchmark the results, we used the Monte Carlo (MC) package TOPAS 3.7 to simulate the generic Ir-192 source recommended by TG186 for MBDCA testing. The source was positioned in the center of a large, cubic water phantom of side 30.1 cm. The phantom contained a boxed-shaped cortical bone heterogeneity of dimensions 2.1x2.1x3.1 cm3, placed with its center 6.0 cm from the source. The total dose was separated into dose from the photon transport generations primary, first scattered (1sc) and msc that all were scored separately with MC, ACE, and ACEcorr. Deviations between ACE and ACEcorr versus MC were analyzed based on the ratio of the dose difference to the total dose, 
, where gen indicates which scatter generation(s) is being tested and where the superscript ACE stands for either ACE or ACEcorr.
Results
The mean dose difference ratio in the cortical bone heterogeneity for ACE was -13±2%, while for the primary, 1sc, and msc dose components it was -0.3±0.4%, 0±1%, and -12±2%, respectively. These results show that the deviation of ACE from the MC system comes from the msc dose calculation. For the corrected algorithm ACEcorr, the mean deviation was reduced to -2±2% for the msc dose. Comparing the calculation times between ACE and ACEcorr showed no significant difference.
Conclusion
We have shown that the dose underestimation to cortical bone by the MBDCA ACE is a consequence of oversimplified multiple scatter transport in non-water media. The results for ACEcorr showed that an improved handling of the balance between local absorption and further transport for multiple scattered photons yields satisfying results for cortical bone without increasing the calculation time.
Supportive Material
Figure 1: A) Color maps of the dose difference ratio to the total dose for ACE (top) and ACEcorr (bottom) versus the MC dose. The left panels show results for the total dose, while the right panels show results for the multiple scatter dose. B) Dose profiles through y=0, normalized to the dose as calculated with the TG-43 formalism (DTG43) for MC (gray solid lines), ACE (gray dash-dotted lines), and ACEcorr (black dotted lines) for the total dose (upper set of graphs) and multiple scatter dose (the lower set of graphs).