Speaker
Description
Specific absorbed fractions (SAF) and S-values (S), which are related to internal radiation dosimetry, can currently be estimated using a variety of Monte Carlo tools, including MCNP and GATE, in order to prevent biological damage from being done to tissues and organs after they've been exposed to ionizing radiation. For physicists with coding skills, such tools make physics easier. However, programming and/or simulation inputs continue to be labor-intensive and time-consuming tasks. In this study, we introduce a newly created Geant4-based code called "DoseCalcs" for internal dosimetry calculations. This code offers a variety of geometrical methods (STL, GDML, TEXT, STL, C++, voxelized, DICOM, and tetrahedral) that can be used to build the simulation geometry, as well as computational capabilities such as running with MPI or multi-threading mode.
The SAFs for eight discrete mono-energetic photons with energies ranging from 0.01 to 2 MeV were estimated using the voxelized ICRP adult female phantom, determined using DoseCalcs, and compared to OpenDose reference data. The accuracy of DoseCalcs is shown to be in good agreement with OpenDose, which shows its suitability for application in the estimation of internal dosimetry quantities using a voxelized geometry method.
While we have parallelized the simulation to a number of sub-simulations with different source configurations on each compute unit, the maximum number of events (G4int size) in a simulation still limits the execution of all source configurations in one execution.
