Program Information
Evaluation of the Electron Condensed History Algorithms in Geant4 for Radiotherapy Simulations
D. J. O'Brien*, G. O. Sawakuchi, The University of Texas MD Anderson Cancer Center, Houston, TX
Presentations
TU-D-205-4 (Tuesday, August 1, 2017) 11:00 AM - 12:15 PM Room: 205
Purpose: To determine the set of condensed history parameters in Geant4 that produces the most consistent results for the Fano cavity test for simulations at radiotherapy energies.
Methods: For each physics list, the Fano cavity test was performed across a range of energies from 0 – 25 MeV. The test was based on that used by Sempau and Andreo (2006) and consisted of an infinite slab geometry of uniform composition (graphite) and density except for a middle layer 2 cm thick which was made of the same material but with a density 1000 times lower than the surrounding medium. To control for the fact that each physics list uses a variety of settings for the various transport parameters (Range Factor, StepLimitType, dRoverRange, Final Range), these tests were also repeated for each case with all of the parameters matching except for the choice of multiple scattering algorithm and ionization model.
Results: Using their default settings, all of the built-in physics lists produced results that were inconsistent with theory with the majority of the results deviating from 3-50% depending on the energy and physics list used. However, when the G4EmStandardPhysicsGS physics list (which uses the Goudsmit-Saunderson multiple scattering model) is used with the StepLimitType set to fUseStep and the Range Factor of 0.02 and Final Range of 0.1 mm, then the results are consistently within 0.1% across the entire energy range studied.
Conclusion: The default parameters for the built-in physics lists do not produce results consistent with theory, however, a physics list employing the Goudsmit-Saunderson multiple scattering algorithm with the right parameters can be used to produce results consistent within 0.1% across the entire therapeutic energy range which is comparable to the consistency of EGSnrc and Penelope.
Funding Support, Disclosures, and Conflict of Interest: This work was partly supported by Elekta.
Contact Email: