Geant4 Model Research Articles

Abstract ID: 194 Validation of GPUMCD for X-ray medical imaging

GPUMCD, a fast GPU-based Monte Carlo transport code, was initially developed and validated for photon and electron dose calculations in radiation oncology. In this work, the code was adapted to X-ray medical imaging applications, notably to simulate Cone-Beam Computed Tomography (CBCT) projections. The photon interaction processes in GPUMCD, initially validated for the therapeutic (megavoltage) energy range, were revisited with imaging and therefore lower energies in mind. Comparisons with physics models available in Geant4 were conducted for the photoelectric effect, Compton and Rayleigh scattering. For Compton scattering, four Geant4 models were used: two from the Geant4 standard physics (G4KleinNishinaModel and G4KleinNishinaCompton), one from Livermore physics (G4LivermoreComptonModel) and one from Penelope physics (G4PenelopeCompton). Three out of the four models consider atomic shells and the energy of the recoil electron. This correction is not considered in G4KleinNishinaCompton and GPUMCD. Tracking of electron was turned off in GPUMCD and Geant4. The photoelectric and Rayleigh scattering algorithms yielded consistent results in GPUMCD and Geant4. However, significant differences were identified between GPUMCD and Geant4 for Compton scattering. More specifically, differences of up to 15% in energy fluence in simulated projections were observed between simulations with and without atomic shell corrections, consistent with the known overestimation of forward scatter in the plain Klein-Nishina model. Implementing shell correction in GPUMCD would significantly increase computation times but appears necessary for imaging applications.

Verification of Electromagnetic Physics Models for Parallel Computing Architectures in the GeantV Project

An intensive R&D and programming effort is required to accomplish new challenges posed by future experimental high-energy particle physics (HEP) programs. The GeantV project aims to narrow the gap between the performance of the existing HEP detector simulation software and the ideal performance achievable, exploiting latest advances in computing technology. The project has developed a particle detector simulation prototype capable of transporting in parallel particles in complex geometries exploiting instruction level microparallelism (SIMD and SIMT), task-level parallelism (multithreading) and high-level parallelism (MPI), leveraging both the multi-core and the many-core opportunities. We present preliminary verification results concerning the electromagnetic (EM) physics models developed for parallel computing architectures within the GeantV project. In order to exploit the potential of vectorization and accelerators and to make the physics model effectively parallelizable, advanced sampling techniques have been implemented and tested. In this paper we introduce a set of automated statistical tests in order to verify the vectorized models by checking their consistency with the corresponding Geant4 models and to validate them against experimental data.

Validation of Physics Models of Geant4 using Data from CMS Experiment

CMS has tuned its simulation program and has chosen a specific physics model of GEANT4 by comparing the simulation results with dedicated test beam experiments. CMS continues to validate the physics models inside GEANT4 using the test beam data as well as collision data. Several physics lists (collection of physics models) inside the most recent version of GEANT4 provide good agreement of the energy response, resolution of pions and protons. The validation results from these studies are used to make the choice of physics list and GEANT4 version for the CMS experiment.

Abstract ID: 36 Geant4 modeling of targeted radionuclide therapy for brain metastasis

Many patients with breast cancer develop brain metastases, which are typically treated through whole-brain irradiation. As a potential alternative treatment, targeted radionuclide therapy (TRT) could be delivered early, targeting the areas of the vasculature where tumor cells are penetrating into the brain. We have developed a Monte Carlo model representing brain vasculature to evaluate and understand a variety of potential therapeutic nuclides: (alpha emitters) Pb-212, At-211, Ac-225, Bi-213, and Tb-149; (beta/Auger electron emitters) Lu-177, Tb-161, I-124, In-111, Y-90, Zr-89, and Ga-67. The micron-scale dose distributions from all radioactive decay products were modeled in Geant4, as well as eV-scale interactions through the G4DNA models [1] . These interactions were then superimposed on an atomic-scale DNA model [2] to estimate strand break yields. Some of the alpha emitters have decay chains with multiple daughter nuclei; we investigate the change in dose profiles and biological effectiveness as a function of time. Alpha emitters have higher doses per decay, and the depth-dose profiles fall off less quickly. The general qualities of the depth-dose profiles are maintained through biologically-relevant variations in vasculature geometry.

Abstract ID: 61 Validation of Geant4 nuclear reaction models for hadrontherapy and preliminary results with SMF and Blob

Reliable nuclear fragmentation models are of utmost importance in hadrontherapy, where Monte Carlo (MC) simulations are used to compute the input parameters of the treatment planning software, to validate the deposited dose calculation, to evaluate the biological effectiveness of the radiation, to correlate the β + emitters production in the patient body with the delivered dose, and to allow a non-invasive treatment verification. Despite of its large use, the models implemented in Geant4 have shown severe limitations in reproducing the measured secondaries yields in ions interaction below 100 MeV/A, in term of production rates, angular and energy distributions [1] , [2] , [3] . We will present a benchmark of the Geant4 models with double-differential cross section and angular distributions of the secondary fragments produced in the 12C fragmentation at 62 MeV/A on thin carbon target, such a benchmark includes the recently implemented model INCL++ [4] , [5] . Moreover, we will present the preliminary results, obtained in simulating the same interaction, with SMF [6] and BLOB [7] . Both, SMF and BLOB are semiclassical one-body approaches to solve the Boltzmann-Langevin equation. They include an identical treatment of the mean-field propagation, on the basis of the same effective interaction, but they differ in the way fluctuations are included. In particular, while SMF employs a Uehling-Uhlenbeck collision term and introduces fluctuations as projected on the density space, BLOB introduces fluctuations in full phase space through a modified collision term where nucleon-nucleon correlations are explicitly involved. Both of them, SMF and BLOB, have been developed to simulate the heavy ion interactions in the Fermi-energy regime. We will show their capabilities in describing 12C fragmentation foreseen their implementation in Geant4.

Abstract ID: 144 An equipment-specific Geant4 model for the Elekta Agility collimator

This study aims to develop and validate a linac-specific Geant4 Monte-Carlo (MC) model of an Elekta Agility collimator (Elekta AB, Sweden) to be used for dose distribution simulations of IMRT and VMAT procedures. The Agility collimator consists of two diaphragms mounted orthogonally to 160 interdigitating leaves. Elekta Precise IAEA phase-space files for 6 MV photon beam were used as primary beam generator. The diaphragms and leaves were primarily modelled following vendor-provided geometry, designed specifically for clinical implementation of MC-based Treatment Planning Systems. However due to small differences in installation and setting, geometric parameters are necessarily linac-specific. Hence such parameters have been defined as tunable mathematical functions, so the user can find the best fit for an individual installation. For model validation, system’s dosimetric characteristics were simulated in Geant4 v.10.01.p02. Fields ranging from 2 × 2 to 20 × 20 cm2 were delivered to a virtual water phantom. Percentage Depth Dose (PDD) and lateral profiles were compared to corresponding measurements conducted with a water phantom (Blue Phantom2, IBA Dosimetry, USA) in an Elekta Synergy (Eleka AB, Sweden) coupled with Elekta Agility. Intraleaf and interleaf leakage and tongue-and-groove (T&G) effect were also investigated. Results show good agreement between measurements and simulations, being comparable to published results [1] . In both cases the maximum leakage and the underdosage due to T&G effect were 0.55% and 20%, respectively. Small discrepancies between measured and simulated PDDs and lateral profiles were observed in small fields, which could be attributed to the use of unsuited phase-spaces. The developed MC model was proven able to reproduce the Elekta Agility collimator. Modelled electron source parameters, as the energy and shape of the incident electron beam, can significantly influence the output of linac dose simulations [2] . A more adequate phase-space for 6 MV photon beam, matching specific configurations of the linac in use, is under development. Validation for IMRT and VMAT plans and for 15 MV photon beam are foreseen.

Neutron transport simulation in high speed moving media using Geant4

A method using Geant4 to simulate neutron transport in moving media is described. The method is implanted in the source code of the software since Geant4 does not intrinsically support a moving object. The simulation utilizes the existing physical model and data library in Geant4, combined with frame transformations to account for the effect of relative velocity between neutrons and the moving media. An example is presented involving a high speed rotating cylinder to verify this method and show the effect of moving media on neutron transport.

Benchmarking GATE/Geant4 for 16O ion beam therapy

Oxygen () ions are a potential alternative to carbon ions in ion beam therapy. Their enhanced linear energy transfer indicates a higher relative biological effectiveness and a reduced oxygen enhancement ratio. Due to the limited availability of ion beams, Monte Carlo (MC) transport codes are important research tools for investigating their potential. The purpose of this study was to validate GATE/Geant4 for ion beam therapy using experimental data from literature. Five hadron physics lists and two electromagnetic options were benchmarked against measured depth dose distributions (DDDs) and charge-changing cross sections. The simulated beam ranges deviated by less than 0.5% for all physics configurations and only a few points exceeded the gamma index criterion (2%/1 mm). However, the simulated partial charge-changing cross sections deviated considerably for some hadron physics configurations. Best agreement with the experimental values was obtained with the quantum molecular dynamics model (QMD), and we therefore suggest using this model in Geant4 to accurately describe the fragmentation of ion beams into lighter fragments ().

Simulation study of neutron production in thick beryllium targets by 35 MeV and 50.5 MeV proton beams

A data-driven nuclear model dedicated to an accurate description of neutron productions in beryllium targets bombarded by proton beams is developed as a custom development that can be used as an add-on to GEANT4 code. The developed model, G4Data(Endf7.1), takes as inputs the total and differential cross section data of ENDF/B-VII.1 for not only the charge-exchange 9Be(p,n)9B reaction which produces discrete neutrons but also the nuclear reactions relevant for the production of continuum neutrons such as 9Be(p,pn)8Be and 9Bep,nα5Li. In our benchmarking simulations for two experiments with 35MeV and 50.5MeV proton beams impinged on 1.16 and 1.05cm thick beryllium targets, respectively, we find that the G4Data(Endf7.1) model can reproduce both the total amounts and the spectral shapes of the measured neutron yield data in a satisfactory manner, while all the considered hadronic models of GEANT4 cannot.

Microdosimetry of electrons in liquid water using the low-energy models of Geant4

The biological effects of ionizing radiation at the cellular level are frequently studied using the well-known formalism of microdosimetry, which provides a quantitative description of the stochastic aspects of energy deposition in irradiated media. Energy deposition can be simulated using Monte Carlo codes, some adopting a computationally efficient condensed-history approach, while others follow a more detailed track-structure approach. In this work, we present the simulation of microdosimetry spectra and related quantities (frequency-mean and dose-mean lineal energies) for incident monoenergetic electrons (50 eV–10 keV) in spheres of liquid water with dimensions comparable to the size of biological targets: base pairs (2 nm diameter), nucleosomes (10 nm), chromatin fibres (30 nm) and chromosomes (300 nm). Simulations are performed using the condensed-history low-energy physics models (“Livermore” and “Penelope”) and the track-structure Geant4-DNA physics models, available in the Geant4 Monte Carlo simulation toolkit. The spectra are compared and the influence of simulation parameters and different physics models, with emphasis on recent developments, is discussed, underlining the suitability of Geant4-DNA models for microdosimetry simulations. It is further shown that with an appropriate choice of simulation parameters, condensed-history transport may yield reasonable results for sphere sizes as small as a few tens of a nanometer.

Lateral variations of radiobiological properties of therapeutic fields of 1H, 4He, 12C and 16O ions studied with Geant4 and microdosimetric kinetic model

As known, in cancer therapy with ion beams the relative biological effectiveness (RBE) of ions changes in the course of their propagation in tissues. Such changes are caused not only by increasing the linear energy transfer (LET) of beam particles with the penetration depth towards the Bragg peak, but also by nuclear reactions induced by beam nuclei leading to the production of various secondary particles. Although the changes of RBE along the beam axis have been studied quite well, much less attention has been paid to the evolution of RBE in the transverse direction, perpendicular to the beam axis. In order to fill this gap, we simulated radiation fields of 1H, 4He, 12C and 16O nuclei of 20 mm in diameter by means of a Geant4-based Monte Carlo model for heavy-ion therapy connected with the modified microdosimetric kinetic model to describe the response of normal ( Gy) and early-responding ( Gy) tissues. Depth and radial distributions of saturation-corrected dose-mean lineal energy, RBE and RBE-weighted dose are investigated for passive beam shaping and active beam scanning. The field of 4He has a small lateral spread as compared with 1H field, and it is characterised by a modest lateral variation of RBE suggesting the use of fixed RBE values across the field transverse cross section at each depth. Reduced uncertainties of RBE on the boundary of a 4He treatment field can be advantageous in a specific case of an organ at risk located in lateral proximity to the target volume. It is found that the lateral distributions of RBE calculated for 12C and 16O fields demonstrate fast variations in the radial direction due to changes of dose and composition of secondary fragments in the field penumbra. Nevertheless, the radiation fields of all four projectiles at radii larger than 20 mm can be characterized by a common RBE value defined by tissue radiosensitivity. These findings can help, in particular, in accessing the transverse homogeneity of radiation fields of ions used in studies in vitro.

New physics model in GEANT4 for the simulation of neutron interactions with organic scintillation detectors

The accurate determination of the response function of organic scintillation neutron detectors complements their experimental characterization. Monte Carlo simulations with GEANT4 can reduce the effort and cost implied, especially for complex detection systems for which the characterization is more challenging. Previous studies have reported on the inaccuracy of GEANT4 in the calculation of the neutron response of organic scintillation detectors above 6 MeV, due to an incomplete description of the neutron-induced alpha production reactions on carbon. We have improved GEANT4 in this direction by incorporating models and data from NRESP, an excellent Monte Carlo simulation tool developed at the Physikalisch-Technische Bundesanstalt (PTB), Germany, for the specific purpose of calculating the neutron response function of organic scintillation detectors. The results have been verified against simulations with NRESP and validated against Time-Of-Flight measurements with an NE213 detector at PTB. This work has potential applications beyond organic scintillation detectors, to other types of detectors where reactions induced by fast neutrons on carbon require an accurate description.

An integrated model of scintillator-reflector properties for advanced simulations of optical transport

Accurately modeling the light transport in scintillation detectors is essential to design new detectors for nuclear medicine or high energy physics. Optical models implemented in software such as Geant4 and GATE suffer from important limitations that we addressed by implementing a new approach in which the crystal reflectance was computed from 3D surface measurements. The reflectance was saved in a look-up-table (LUT) then used in Monte Carlo simulation to determine the fate of optical photons. Our previous work using this approach demonstrated excellent agreement with experimental characterization of crystal light output in a limited configuration, i.e. when using no reflector.As scintillators are generally encapsulated in a reflector, it is essential to include the crystal-reflector interface in the LUT. Here we develop a new LUT computation and apply it to several reflector types. A second LUT that contains transmittance data is also saved to enable modeling of optical crosstalk.LUTs have been computed for rough and polished crystals coupled to a Lambertian (e.g. Teflon tape) or a specular reflector (e.g. ESR) using air or optical grease, and the light output was computed using a custom Monte Carlo code. 3 × 3 × 20 mm3 lutetium oxyorthosilicate crystals were prepared using these combinations, and the light output was measured experimentally at different irradiation depths. For all reflector and surface finish combinations, the measured and simulated light output showed very good agreement.The behavior of optical photons at the interface crystal-reflector was studied using these simulations, and results highlighted the large difference in optical properties between rough and polished crystals, and Lambertian and specular reflectors. These simulations also showed how the travel path of individual scintillation photons was affected by the reflector and surface finish.The ultimate goal of this work is to implement this model in Geant4 and GATE, and provide a database of scintillators combined with a variety of reflectors.

Backscattered electron emission after proton impact on carbon and gold films: Experiments and simulations

This work aims at measuring the proton induced secondary electron energy spectra from nanometer thin films. Backscattered electron energy spectra were measured within an energy range from 0 to 600eV using a Retarding Field Analyser (RFA). This paper presents energy spectra obtained for proton (0.5MeV; 1MeV; 1.5MeV; 2MeV) irradiation of thin carbon films (50 and 100nm thick) and thin gold film (200nm). These experimental spectra were compared with Monte Carlo simulations based on TRAX code and Geant4 simulation toolkit. Good agreement between experimental, TRAX and Geant4 results were observed for the carbon target. For the gold target, we report major differences between both Monte Carlo environments. Limitation of Geant4 models for low energy electron emission was highlighted. On the contrary, TRAX simulations present encouraging results for the modeling of low-energy electron emission from gold target.

Measurement of antiproton annihilation on Cu, Ag and Au with emulsion films

The characteristics of low energy antiproton annihilations on nuclei (e.g. hadronization and product multiplicities) are not well known, and Monte Carlo simulation packages that use different models provide different descriptions of the annihilation events. In this study, we measured the particle multiplicities resulting from antiproton annihilations on nuclei. The results were compared with predictions obtained using different models in the simulation tools GEANT4 and FLUKA. For this study, we exposed thin targets (Cu, Ag and Au) to a very low energy antiproton beam from CERN's Antiproton Decelerator, exploiting the secondary beamline available in the AEgIS experimental zone. The antiproton annihilation products were detected using emulsion films developed at the Laboratory of High Energy Physics in Bern, where they were analysed at the automatic microscope facility. The fragment multiplicity measured in this study is in good agreement with results obtained with FLUKA simulations for both minimally and heavily ionizing particles.

A comparison between track-structure, condensed-history Monte Carlo simulations and MIRD cellular S-values

The S-value is a standard measure in cellular dosimetry. S-values are calculated by applying analytical methods or by Monte Carlo simulation. In Monte Carlo simulation, particles are either tracked individually event-by-event or close events are condensed and processed collectively in different steps. Both of these methods have been employed for estimation of cellular S-values, but there is no consistency between the published results. In the present paper, we used the Geant4-DNA track-structure physics model as the reference to estimate the cellular S-values. We compared the results with the corresponding values obtained from the following three condensed-history physics models of Geant4: Penelope, Livermore and standard. The geometry and source were exactly the same in all the simulations. We utilized mono-energetic electrons with an initial kinetic energy in the range 1–700 keV as the source of radiation. We also compared our results with the MIRD S-values. We first drew an overall comparison between different data series and then compared the dependence of results on the energy of particles and the size of scoring compartments. The overall comparison indicated a very good linear correlation (R2 > 91%) and small bias (3%) between the results of the track-structure model and the condensed-history physics model. The bias between MIRD and the results of Monte Carlo track-structure simulation was considerable (−8%). However, the point-by-point comparison revealed differences of up to 28% between the condensed-history and the track-structure MC codes for self-absorption S-values in the 10–50 keV energy range. For the cross-absorption S-values, the difference was up to 34%. In this energy range, the difference between the MIRD S-values and the Geant4-DNA results was up to 68%. Our findings suggest that the consistency/inconsistency of the results obtained with different MC simulations depends on the size of the scoring volumes, the energy of the particles, the step-size in electron tracking and the energy cutoff used in the MC codes.

Validation of Geant4 on Proton Transportation for Thick Absorbers: Study Based on Tschalär Experimental Data

Imaging techniques using protons as incident particles are currently being developed to substitute X-ray computer tomography and nuclear magnetic resonance methods in proton therapy. They deal with relatively thick targets, like the human head or trunk, where protons lose a significant part of their energy, however, they have enough energy to exit the target. The physical quantities important in proton imaging are kinetic energy, angle and coordinates of emerging proton from an absorber material. In recent times, many research groups use the Geant4 toolkit to simulate proton imaging devices. Most of the available publications about validation of Geant4 models are for thin or thick absorbers (Bragg Peak studies), that are not consistent with the contour conditions applied to proton imaging. The main objective of this work is to evaluate the kinetic energy spectrum for protons emerging from homogeneous absorbers slabs comparing it to the experimental results published by Tschalar and Maccabee, in 1970. Different models (standard and detailed) available on Geant4 (version 9.6.p03) are explored taking into account its accuracy and computational performance. This paper presents a validation for protons with incident kinetic energies of 19.68 MeV and 49.10 MeV. The validation results from the emerging protons kinetic energy spectra show that: (i) there are differences between the reference data and the data produced by different processes evoked for transportation and (ii) the validation energies are sensitive to sub-shell processes.

ELECTRON ABSORBED FRACTIONS IN AN IMAGE-BASED MICROSCOPIC SKELETAL DOSIMETRY MODEL OF CHINESE ADULT MALE.

Based on the Chinese reference adult male voxel model, a set of microscopic skeletal models of Chinese adult male is constructed through the processes of computed tomography (CT) imaging, bone coring, micro-CT imaging, image segmentation, merging into macroscopic bone model and implementation in Geant4. At the step of image segmentation, a new bone endosteum (BE) segmentation method is realized by sampling. The set of model contains 32 spongiosa samples with voxel size of 19 μm cubes. The microscopic spongiosa bone data for Chinese adult male are provided. Electron absorbed fractions in red bone marrow (RBM) and BE are calculated. Source tissues include the bone marrow (red and yellow), trabecular bone (surfaces and volumes) and cortical bone (surfaces and volumes). Target tissues include RBM and BE. Electron energies range from 10 keV to 10 MeV. Additionally, comparison of the result with other investigations is provided.

A computationally effcient moment-preserving Monte Carlo proton transport method in Geant4

The moment-preserving method, demonstrated as a viable alternative to condensed history for electrons, is extended to protons. Given the generality and the flexibility of the method, a discrete Coulomb scattering and discrete impactionization differential cross-section library for protons was readily developed and existing Geant4 electron discrete process and model classes were extended to make use of the new proton library. It is shown that levels of effciency and accuracy similar to those demonstrated for electrons are obtainable for protons in straight-ahead, energy-loss problems. However, in problems with deflection, agreement is strongly dependent on the proton energy. That is, good agreement was observed in the few MeV range, while unsatisfactory agreement was observed in problems with proton energies above 100-MeV.

Inverse-collimated proton radiography for imaging thin materials.

Relativistic, magnetically focused proton radiography was invented at Los Alamos National Laboratory using the 800 MeV LANSCE beam and is inherently well-suited to imaging dense objects, at areal densities >20 g cm-2. However, if the unscattered portion of the transmitted beam is removed at the Fourier plane through inverse-collimation, this system becomes highly sensitive to very thin media, of areal densities <100 mg cm-2. Here, this inverse-collimation scheme is described in detail and demonstrated by imaging Xe gas with a shockwave generated by an aluminum plate compressing the gas at Mach 8.8. With a 5-mrad inverse collimator, an areal density change of just 49 mg cm-2 across the shock front is discernible with a contrast-to-noise ratio of 3. Geant4 modeling of idealized and realistic proton transports can guide the design of inverse-collimators optimized for specific experimental conditions and show that this technique performs better for thin targets with reduced incident proton beam emittance. This work increases the range of areal densities to which the system is sensitive to span from ∼25 mg cm-2 to 100 g cm-2, exceeding three orders of magnitude. This enables the simultaneous imaging of a dense system as well as thin jets and ejecta material that are otherwise difficult to characterize with high-energy proton radiography.