Photon Transport Simulation Research Articles

Graphics processing units-accelerated adaptive nonlocal means filter for denoising three-dimensional Monte Carlo photon transport simulations.

.The Monte Carlo (MC) method is widely recognized as the gold standard for modeling light propagation inside turbid media. Due to the stochastic nature of this method, MC simulations suffer from inherent stochastic noise. Launching large numbers of photons can reduce noise but results in significantly greater computation times, even with graphics processing units (GPU)-based acceleration. We develop a GPU-accelerated adaptive nonlocal means (ANLM) filter to denoise MC simulation outputs. This filter can effectively suppress the spatially varying stochastic noise present in low-photon MC simulations and improve the image signal-to-noise ratio (SNR) by over 5 dB. This is equivalent to the SNR improvement of running nearly more photons. We validate this denoising approach using both homogeneous and heterogeneous domains at various photon counts. The ability to preserve rapid optical fluence changes is also demonstrated using domains with inclusions. We demonstrate that this GPU-ANLM filter can shorten simulation runtimes in most photon counts and domain settings even combined with our highly accelerated GPU MC simulations. We also compare this GPU-ANLM filter with the CPU version and report a threefold to fourfold speedup. The developed GPU-ANLM filter not only can enhance three-dimensional MC photon simulation results but also be a valuable tool for noise reduction in other volumetric images such as MRI and CT scans.

Monte Carlo Simulation of Photon Transport for Computing Fluence Rate in Biological Tissue

Monte Carlo Simulation of Photon Transport for Computing Fluence Rate in Biological Tissue

Monte Carlo Simulation of Photon Transport for Computing Fluence Rate in Biological Tissue

In this research, we applied the Monte Carlo method to simulate photon transport in a biological tissue, consisting of epidermal, dermis, and blood vessels. Particularly, we computed the fluence rate of the system at a wavelength of 400 nm using different beam sources, such as collimated beam, Gaussian beam, and isotropic point sources. In addition, the fluence rate is calculated within the collimated beam at different wavelengths between 300–1000 nm by considering the absorption coefficient (μa) for blood, dermis, and epidermis. For the collimated beam, the resulting fluence rate was found almost similar in the case of the epidermis and dermis at wavelengths between 600–1000 nm, whereas the blood vessels occur at a wavelength of 400 nm with a maximum absorption coefficient of blood (μa) of 3586 cm–1. The present study illustrated the ability of the penetration of light in biological tissues and the escaped light could provide the information about the components of the biological tissue. From the theoretical perspective, the comprehension of light-tissue interactions can support the field of biomedical optics.

Optical photon transport and geometry contributions to time response of scintillation detectors

The sensitivities of both time and energy resolutions of a typical scintillation detector to major optical parameters (i.e. paint reflectivity, quantum efficiency of photomultiplier tube and attenuation coefficient) have been estimated using a dedicated Monte Carlo (MC) optical photon transport (OPT) simulation code, PHOTRACK, and the OPT capabilities of general-purpose code, FLUKA. Both cylindrical and parallelepiped geometries have been considered for the scintillator cell and lightguide. The results determine the scintillation light wavelength regions that the energy and time resolution represent enhanced sensitivities to small change/uncertainty in optical parameters.

Re-evaluation of pediatric 18F-FDG dosimetry: Cristy–Eckerman versus UF/NCI hybrid computational phantoms

Because of the concerns associated with radiation exposure at a young age, there is an increased interest in pediatric absorbed dose estimates for imaging agents. Almost all reported pediatric absorbed dose estimates, however, have been determined using adult pharmacokinetic data with radionuclide S values that take into account the anatomical differences between adults and children based upon the older Cristy–Eckerman (C–E) stylized phantoms. In this work, we use pediatric model-derived pharmacokinetics to compare absorbed dose and effective dose estimates for 18F-FDG in pediatric patients using S values generated from two different geometries of computational phantoms. Time-integrated activity coefficients of 18F-FDG in brain, lungs, heart wall, kidneys and liver, retrospectively, calculated from 35 pediatric patients at the Boston’s Children Hospital were used. The absorbed dose calculation was performed in accordance with the Medical Internal Radiation Dose method using S values generated from the University of Florida/National Cancer Institute (UF/NCI) hybrid phantoms, as well as those from C–E stylized computational phantoms. The effective dose was computed using tissue-weighting factors from ICRP Publication 60 and ICRP Publication 103 for the C–E and UF/NCI, respectively. Substantial differences in the absorbed dose estimates between UF/NCI hybrid pediatric phantoms and the C–E stylized phantoms were found for the lungs, ovaries, red bone marrow and urinary bladder wall. Large discrepancies in the calculated dose values were observed in the bone marrow; ranging between −26% to +199%. The effective doses computed by the UF/NCI hybrid phantom S values were slightly different than those seen using the C–E stylized phantoms with percent differences of −0.7%, 2.9% and 2.5% for a newborn, 1 year old and 5 year old, respectively. Differences in anatomical modeling features among computational phantoms used to perform Monte Carlo-based photon and electron transport simulations for 18F, and very likely for other radionuclides, impact internal organ dosimetry computations for pediatric nuclear medicine studies.

High-accuracy Geant4 simulation and semi-analytical modeling of nuclear resonance fluorescence

Nuclear resonance fluorescence (NRF) is a photonuclear interaction that enables highly isotope-specific measurements in both pure and applied physics scenarios. High-accuracy design and analysis of NRF measurements in complex geometries is aided by Monte Carlo simulations of photon physics and transport, motivating Jordan and Warren (2007) to develop the G4NRF codebase for NRF simulation in Geant4. In this work, we enhance the physics accuracy of the G4NRF code and perform improved benchmarking simulations. The NRF cross section calculation in G4NRF, previously a Gaussian approximation, has been replaced with a full numerical integration for improved accuracy in thick-target scenarios. A high-accuracy semi-analytical model of expected NRF count rates in a typical NRF measurement is then constructed and compared against G4NRF simulations for both simple homogeneous and more complex heterogeneous geometries. Agreement between rates predicted by the semi-analytical model and G4NRF simulation is found at a level of ${\sim}1\%$ in simple test cases and ${\sim}3\%$ in more realistic scenarios, improving upon the ${\sim}20\%$ level of the initial benchmarking study and establishing a highly-accurate NRF framework for Geant4.

The Failure of Monte Carlo Radiative Transfer at Medium to High Optical Depths

Abstract Computer simulations of photon transport through an absorbing and/or scattering medium form an important research tool in astrophysics. Nearly all software codes performing such simulations for three-dimensional geometries employ the Monte Carlo (MC) radiative transfer (RT) method, including various forms of biasing to accelerate the calculations. Because of the probabilistic nature of the MC technique, the outputs are inherently noisy, but it is often assumed that the average values provide the physically correct result. We show that this assumption is not always justified. Specifically, we study the intensity of radiation penetrating an infinite, uniform slab of material that absorbs and scatters the radiation with equal probability. The basic MCRT method, without any biasing mechanisms, starts to break down for transverse optical depths τ ≳ 20 because so few of the simulated photon packets reach the other side of the slab. When including biasing techniques such as absorption/scattering splitting and path length stretching, the simulated photon packets do reach the other side of the slab but the biased weights do not necessarily add up to the correct solution. While the noise levels seem to be acceptable, the average values sometimes severely underestimate the correct solution. Detecting these anomalies requires the judicious application of statistical tests, similar to those used in the field of nuclear particle transport, possibly in combination with convergence tests employing consecutively larger numbers of photon packets. In any case, for transverse optical depths τ ≳ 75 the MC methods used in our study fail to solve the one-dimensional slab problem, implying the need for approximations such as a modified random walk.

Multi-threading performance of Geant4, MCNP6, and PHITS Monte Carlo codes for tetrahedral-mesh geometry

In this study, the multi-threading performance of the Geant4, MCNP6, and PHITS codes was evaluated as a function of the number of threads (N) and the complexity of the tetrahedral-mesh phantom. For this, three tetrahedral-mesh phantoms of varying complexity (simple, moderately complex, and highly complex) were prepared and implemented in the three different Monte Carlo codes, in photon and neutron transport simulations. Subsequently, for each case, the initialization time, calculation time, and memory usage were measured as a function of the number of threads used in the simulation. It was found that for all codes, the initialization time significantly increased with the complexity of the phantom, but not with the number of threads. Geant4 exhibited much longer initialization time than the other codes, especially for the complex phantom (MRCP). The improvement of computation speed due to the use of a multi-threaded code was calculated as the speed-up factor, the ratio of the computation speed on a multi-threaded code to the computation speed on a single-threaded code. Geant4 showed the best multi-threading performance among the codes considered in this study, with the speed-up factor almost linearly increasing with the number of threads, reaching ~30 when N = 40. PHITS and MCNP6 showed a much smaller increase of the speed-up factor with the number of threads. For PHITS, the speed-up factors were low when N = 40. For MCNP6, the increase of the speed-up factors was better, but they were still less than ~10 when N = 40. As for memory usage, Geant4 was found to use more memory than the other codes. In addition, compared to that of the other codes, the memory usage of Geant4 more rapidly increased with the number of threads, reaching as high as ~74 GB when N = 40 for the complex phantom (MRCP). It is notable that compared to that of the other codes, the memory usage of PHITS was much lower, regardless of both the complexity of the phantom and the number of threads, hardly increasing with the number of threads for the MRCP.

Simulated and measured neutron/gamma light output distribution for poly-energetic neutron/gamma sources

In the present paper, the light output distribution due to poly-energetic neutron/gamma (neutron or gamma) source was calculated using the developed MCNPX-ESUT-PE (MCNPX-Energy engineering of Sharif University of Technology-Poly Energetic version) computational code. The simulation of light output distribution includes the modeling of the particle transport, the calculation of scintillation photons induced by charged particles, simulation of the scintillation photon transport and considering the light resolution obtained from the experiment. The developed computational code is able to simulate the light output distribution due to any neutron/gamma source. In the experimental step of the present study, the neutron-gamma discrimination based on the light output distribution was performed using the zero crossing method. As a case study, 241Am–9Be source was considered and the simulated and measured neutron/gamma light output distributions were compared. There is an acceptable agreement between the discriminated neutron/gamma light output distributions obtained from the simulation and experiment.

FNIRS Optodes\u2019 Location Decider (fOLD): a toolbox for probe arrangement guided by brain regions-of-interest

The employment of functional near-infrared spectroscopy (fNIRS) as a method of brain imaging has increased over the last few years due to its portability, low-cost and robustness to subject movement. Experiments with fNIRS are designed in the face of a limited number of sources and detectors (optodes) to be positioned on selected portion(s) of the scalp. The optodes locations represent an expectation of assessing cortical regions relevant to the experiment’s hypothesis. However, this translation process remains a challenge for fNIRS experimental design. In the present study, we propose an approach that automatically decides the location of fNIRS optodes from a set of predefined positions with the aim of maximizing the anatomical specificity to brain regions-of-interest. The implemented method is based on photon transport simulations on two head atlases. The results are compiled into the publicly available “fNIRS Optodes’ Location Decider” (fOLD). This toolbox is a first-order approach to bring the achieved advancements of parcellation methods and meta-analyses from functional magnetic resonance imaging to more precisely guide the selection of optode positions for fNIRS experiments.

Scalable and massively parallel Monte Carlo photon transport simulations for heterogeneous computing platforms.

.We present a highly scalable Monte Carlo (MC) three-dimensional photon transport simulation platform designed for heterogeneous computing systems. Through the development of a massively parallel MC algorithm using the Open Computing Language framework, this research extends our existing graphics processing unit (GPU)-accelerated MC technique to a highly scalable vendor-independent heterogeneous computing environment, achieving significantly improved performance and software portability. A number of parallel computing techniques are investigated to achieve portable performance over a wide range of computing hardware. Furthermore, multiple thread-level and device-level load-balancing strategies are developed to obtain efficient simulations using multiple central processing units and GPUs.

Electroluminescent cooling in intracavity light emitters: modeling and experiments

We develop a coupled electronic charge and photon transport simulation model to allow for deeper analysis of our recent experimental studies of intracavity double diode structures (DDSs). The studied structures consist of optically coupled AlGaAs/GaAs double heterojunction light emitting diode (LED) and GaAs p–n-homojunction photodiode (PD) structure, integrated as a single semiconductor device. The drift–diffusion formalism for charge transport and an optical model, coupling the LED and the PD, are self-consistently applied to complement our experimental work on the evaluation of the efficiency of these DDSs. This is to understand better their suitability for electroluminescent cooling (ELC) demonstration, and shed further light on electroluminescence and optical energy transfer in the structures. The presented results emphasize the adverse effect of non-radiative recombination on device efficiency, which is the main obstacle for achieving ELC in III-V semiconductors.

Abstract ID: 69 Validation of XRMC Monte Carlo toolkit for external dosimetry applied to mammography

The objective of this project is to validate for macroscopic quantities the xrmc toolkit version 6.5.0-2 [1] , comparing against experimental data collected for dosimetric applications in mammography (external dosimetry). The xrmc is focused on X-ray photon transport simulation, and makes extensive use of variance reduction techniques. To validate the simulated quantities a database was generated/organized with experimental data collected from three different mammographic equipments (Mammomat Inspiration, Mammomat 3000 and LORAD MIII). The KERMA, exposure, half-value-layer (HVL) and backscattering (BSF) were the quantities used in the study. The evoked modes on simulation transport for xrmc toolkit were: transmission mode and mode with scattering. This is an ongoing work and at the moment the data analyzes have been partially performed. To evaluate the differences in results due the differences on description of the incident spectra it was modeled spectra from two different resources: catalogue IPEM Report 78 [2] and the spectra generated to Lorad MIII equipment based on model published by Tucker et al. 1991 [3] . The preliminary results show good agreement between xrmc and experimental data for HVL. We are planning to present the complete set of absolute and derived quantities comparing each transport model. This is a preliminary evaluation of the toolkit where its interface and its options in the devices are being tested and few changes in the code are being defined to make xrmc more user friendly and more adequate to simulate external dosimetry in mammography. Research developed with the partial support of the National Supercomputing Center, Federal University of Rio Grande do Sul.

Skin color simulation - review and analysis of available Monte Carlo-based photon transport simulation models

Skin color simulation - review and analysis of available Monte Carlo-based photon transport simulation models

Voxelized ray-tracing simulation dedicated to multi-pinhole molecular breast tomosynthesis

Accurate gamma photon transport simulations of emission tomography systems are important to optimise system geometries and for iterative image reconstruction. Monte Carlo simulation (MCS) is widely established for this purpose but has the disadvantage of being prohibitively slow. Voxelized ray tracing (VRT) can be used as an alternative but the accuracy of VRT needs to be assessed for each simulation task at hand. The aim of this work is to propose and validate a dedicated VRT code for a novel radionuclide-based multi-pinhole molecular breast tomosynthesis (MP-MBT) scanner. The MP-MBT system images radionuclide distributions in a mildly compressed breast using two opposing gamma cameras, each equipped with a focusing multi-pinhole collimator, that slide along opposite sides of the breast. VRT simulates gamma photon transport by tracing rays efficiently through the voxelized phantom, collimator, and detector volumes using Siddon’s raytracing algorithm, accelerated by dual-grid methods. To assess its accuracy, we compare point spread functions (PSFs) calculated with VRT for different voxel sizes with those generated by the established MCS toolkit GATE. Furthermore, VRT and MCS-simulated projections of realistic anthropomorphic XCAT phantoms with different compressed breast sizes are compared, as well as reconstructed images obtained from these projections. With VRT, PSFs for MP-MBT can be simulated accurately when the fine voxel size of the VRT’s dual-grid is 1/8 mm. Reaching a similar deviation from noiseless PSFs takes 29 300 times longer with full MCS than with VRT. Furthermore, XCAT phantom simulations show that VRT-generated projections are very close to MCS-generated low-noise projections when these are corrected for scatter by the triple energy window method. However, we also find that primary gamma photons from the torso may in some cases reach the detector, meaning that torso activity should not be neglected in VRT. Finally, reconstructed images obtained from projections generated by VRT and MCS are visually very similar and have no significant difference in contrast and noise characteristics. We conclude that VRT can accurately and efficiently simulate MP-MBT even though it neglects scattered photons originating from the torso.

A Monte-Carlo Benchmark of TRIPOLI-4® and MCNP on ITER neutronics

Radiation protection and shielding studies are often based on the extensive use of 3D Monte-Carlo neutron and photon transport simulations. ITER organization hence recommends the use of MCNP-5 code (version 1.60), in association with the FENDL-2.1 neutron cross section data library, specifically dedicated to fusion applications. The MCNP reference model of the ITER tokamak, the ‘C-lite’, is being continuously developed and improved. This article proposes to develop an alternative model, equivalent to the 'C-lite', but for the Monte-Carlo code TRIPOLI-4®. A benchmark study is defined to test this new model. Since one of the most critical areas for ITER neutronics analysis concerns the assessment of radiation levels and Shutdown Dose Rates (SDDR) behind the Equatorial Port Plugs (EPP), the benchmark is conducted to compare the neutron flux through the EPP. This problem is quite challenging with regard to the complex geometry and considering the important neutron flux attenuation ranging from 1014 down to 108 n•cm-2 •s-1 . Such code-to-code comparison provides independent validation of the Monte-Carlo simulations, improving the confidence in neutronic results.

Geant4 simulation of optical photon transport in scintillator tile with direct readout by silicon photomultiplier

The direct coupling of silicon photomultiplier to the scintillator tile is considered to be the main option for active elements of the highly granular hadron calorimeter developed for future linear collider experiments. In this study, the response of the scintillator-SiPM system to minimum ionising particles was simulated using the optical photon transport functionality available in the Geant4 package. The uniformity of response for both flat tile and tile with dimple was estimated from the simulations and compared to the experimental results obtained in the previous studies.

A study of the radiative transfer equation using a spherical harmonics-nodal collocation method

Optical tomography has found many medical applications that need to know how the photons interact with the different tissues. The majority of the photon transport simulations are done using the diffusion approximation, but this approximation has a limited validity when optical properties of the different tissues present large gradients, when structures near the photons source are studied or when anisotropic scattering has to be taken into account. As an alternative to the diffusion model, the PL equations for the radiative transfer problem are studied. These equations are discretized in a rectangular mesh using a nodal collocation method. The performance of this model is studied by solving different 1D and 2D benchmark problems of light propagation in tissue having media with isotropic and anisotropic scattering.

Monte Carlo modeling of a novel depth-encoding PET detector with DETECT2000

We designed a depth-encoding positron emission tomography (PET) detector using wavelength shifting (WLS) fibers for a low-cost and high-resolution PET. In order to characterize the theoretical performance of the newly-designed PET detector, we performed a DETECT2000 simulation of the optical photon transport in the crystal array and the WLS fibers. The detector module is made of an array of individual crystals, WLS fibers, and multi-pixel photon counters (MPPCs). The lutetium oxyorthosilicate (LSO) crystal array consists of 15 × 15 crystals with a 2 mm × 2 mm surface size and three different heights: 18, 20, and 22 mm. On every other line, scintillators 2 mm shorter in height are arranged in the y-direction on the top and the x-direction on the bottom of the array to make grooves for insertion of a WLS fiber. The total size of the detector module is 30 mm × 30 mm × 22 mm. One end of each WLS fiber is coupled to MPPCs to measure scintillating light trapped in the fiber. The depth-of-interaction (DOI) resolution and pixel identification capability are determined by using a DETECT2000 simulation. All pixels of the 15 × 15 LSO array are well decoded, and the DOI resolution ranges from 1.43 mm to 2.53 mm with an average of 2.08 mm full width at half maximum (FWHM). These results indicate that the proposed detector performs respectably for high-resolution PET scanning.

Modeling parameterized geometry in GPU-based Monte Carlo particle transport simulation for radiotherapy

Monte Carlo (MC) particle transport simulation on a graphics-processing unit (GPU) platform has been extensively studied recently due to the efficiency advantage achieved via massive parallelization. Almost all of the existing GPU-based MC packages were developed for voxelized geometry. This limited application scope of these packages. The purpose of this paper is to develop a module to model parametric geometry and integrate it in GPU-based MC simulations. In our module, each continuous region was defined by its bounding surfaces that were parameterized by quadratic functions. Particle navigation functions in this geometry were developed. The module was incorporated to two previously developed GPU-based MC packages and was tested in two example problems: (1) low energy photon transport simulation in a brachytherapy case with a shielded cylinder applicator and (2) MeV coupled photon/electron transport simulation in a phantom containing several inserts of different shapes. In both cases, the calculated dose distributions agreed well with those calculated in the corresponding voxelized geometry. The averaged dose differences were 1.03% and 0.29%, respectively. We also used the developed package to perform simulations of a Varian VS 2000 brachytherapy source and generated a phase-space file. The computation time under the parameterized geometry depended on the memory location storing the geometry data. When the data was stored in GPU’s shared memory, the highest computational speed was achieved. Incorporation of parameterized geometry yielded a computation time that was ~3 times of that in the corresponding voxelized geometry. We also developed a strategy to use an auxiliary index array to reduce frequency of geometry calculations and hence improve efficiency. With this strategy, the computational time ranged in 1.75–2.03 times of the voxelized geometry for coupled photon/electron transport depending on the voxel dimension of the auxiliary index array, and in 0.69–1.23 times for photon only transport.