A comparison between track-structure, condensed-history Monte Carlo simulations and MIRD cellular S-values
M A Tajik-Mansoury1, H Rajabi2, H Mozdarani3
1Faculty of Medical Sciences, Department of Medical Physics, Semnan University of Medical Sciences, Semnan, Iran.
Physics in Medicine and Biology
|February 10, 2017
Summary
Monte Carlo simulations for cellular S-values show good agreement between track-structure and condensed-history models, but significant differences with MIRD values, especially at lower energies.
Area of Science:
- Biophysics and Radiation Dosimetry
- Computational Physics and Modeling
Background:
- Cellular S-values are crucial for dosimetry, calculated using analytical methods or Monte Carlo (MC) simulations.
- Existing MC methods for S-value estimation lack consistency, necessitating comparative studies.
- Geant4-DNA's track-structure model offers a detailed approach to particle transport simulation.
Purpose of the Study:
- To evaluate and compare cellular S-values estimated using the Geant4-DNA track-structure model against Geant4's condensed-history physics models (Penelope, Livermore, standard).
- To assess the consistency of MC simulation results with established MIRD S-values.
- To investigate the influence of particle energy and scoring compartment size on S-value discrepancies.
Main Methods:
- Utilized the Geant4-DNA track-structure physics model as a reference for cellular S-value calculations.
- Compared results with three Geant4 condensed-history models (Penelope, Livermore, standard) using identical geometry and sources.
- Employed mono-energetic electrons (1-700 keV) and compared findings with MIRD S-values.
Main Results:
- Overall, a strong linear correlation (R² > 91%) and low bias (3%) were observed between track-structure and condensed-history models.
- A considerable bias (-8%) was found between MIRD values and the MC track-structure simulation results.
- Point-by-point comparisons revealed up to 28% differences for self-absorption and 34% for cross-absorption S-values between condensed-history and track-structure models in the 10-50 keV range.
Conclusions:
- The Geant4-DNA track-structure model provides a reliable reference for cellular S-value estimation.
- Discrepancies between MC simulation methods and MIRD values are significant, particularly at lower electron energies.
- The consistency of MC S-value results is sensitive to scoring volume size, particle energy, tracking step-size, and energy cutoffs.
