Summary
Highlights
The main goal of radiotherapy is to induce mitotic death or apoptosis in malignant tumor cells while minimizing damage to healthy tissues. This involves both indirect action (electron interaction with water to form radicals) and direct action (electrons directly interacting with target molecules). Accuracy is critical, as errors exceeding 5% can lead to significant alterations in tumor control probability and normal tissue complication probabilities, increasing the risk of secondary cancers.
Radiotherapy treatment typically involves CT scanning for localization, skin reference mark treatment planning, virtual simulation, and actual treatment. Clinical algorithms are needed to simulate these procedures. While simple scatter is an older method, model-based approaches like convolution superposition are more advanced. However, most current algorithms struggle to accurately predict dose distribution in small fields where lateral equilibrium is not achieved.
Geant4 (Geometry and Tracking) is a C++ programming toolkit designed for simulating the passage of particles through matter. It is well-validated, offers sophisticated geometry description, and can handle both Boolean and analytical phantoms. The objective of this research is to reproduce a realistic clinical beam for treatment planning and model the internal structure of an Electra compact linear accelerator. This provides practical guidance on using Geant4, offering a cost-effective alternative to actual dosimetry experiments.
The methodology involves both experimental and simulation phases. Simulation requirements include defining geometry and material composition, source definition, physics model/interaction data table, a random number generator, and a detector. The computing platform used includes a 3.07 GHz processor, GCC 4.1.2 compiler, and the PC and Midlineup package. The head assembly of the Electra linear accelerator was modeled, with tungsten or tungsten alloy being key material compositions.
The simulation involves bombardment with 2 billion electrons, production of PSF (Phase Space File) before the upper jaw, application of PSF in varying field line sizes, and evaluation of dose distributions. Visualization during interaction time and PSF generation for different primary events are key steps. Simulation efficiency shows a rate of 238.7 particles per second with visualization on, and 2006.5 particles per second with visualization off for 1 million particles. The simulated data represents dose versus depth size.
Validation involved experimental procedures using the Electra compact linear accelerator, a PTW MP3 water phantom, a dual-channel electrometer, and an ionization chamber. An error estimator introduced by Giglio et al. (2011) was used for comparison. The relative PDD (Percent Depth Dose) versus depth in millimeters was compared between Monte Carlo simulation and actual experiment for varying field sizes. Calculated errors were 4.6% for a 10x10 square cm field and 3.9% for a 15x15 square cm field.
Geant4 successfully modeled complex geometries of the Electra compact linear accelerator, demonstrating good agreement between simulated and measured beam data along the central axis. The developed codes show potential for radiotherapy applications. Recommendations include running simulations on higher computing platforms, simulating beam profiles, and fine-tuning beam divergence.