Table of contents

Monte Carlo particle transport techniques offer exciting tools for radiotherapy research, where they play an increasingly important role. Topics of research related to clinical applications range from treatment planning, motion and registration studies, brachytherapy, verification imaging and dosimetry.

The International Workshop on Monte Carlo Techniques in Radiotherapy Delivery and Verification took place in a hotel in Montreal in French Canada, from 29 May–1 June 2007, and was the third workshop to be held on a related topic, which now seems to have become a tri-annual event. About one hundred workers from many different countries participated in the four-day meeting. Seventeen experts in the field were invited to review topics and present their latest work. About half of the audience was made up by young graduate students. In a very full program, 57 papers were presented and 10 posters were on display during most of the meeting. On the evening of the third day a boat trip around the island of Montreal allowed participants to enjoy the city views, and to sample the local cuisine.

The topics covered at the workshop included the latest developments in the most popular Monte Carlo transport algorithms, fast Monte Carlo, statistical issues, source modeling, MC treatment planning, modeling of imaging devices for treatment verification, registration and deformation of images and a sizeable number of contributions on brachytherapy.

In this volume you will find 27 short papers resulting from the workshop on a variety of topics, some of them on very new stuff such as graphics processing units for fast computing, PET modeling, dual-energy CT, calculations in dynamic phantoms, tomotherapy devices, . . . .

We acknowledge the financial support of the National Cancer Institute of Canada, the Institute of Cancer Research of the Canadian Institutes of Health Research, the Association Québécoise des Physicien(ne)s Médicaux Clinique, the Institute of Physics, and MedicalPhysicsWeb. At McGill we thank the following departments for support: the Cancer Axis of the Research Institute of the McGill University Health Center, the Faculties of Medicine and Science, the Departments of Oncology and Physics and the Medical Physics Unit. The following companies are thanked: TomoTherapy and Standard Imaging. The American Association of Physicists in Medicine and the International Atomic Energy Agency are gratefully acknowledged for endorsing the meeting.

A final word of thanks goes out to all of those who contributed to the successful Workshop: first of all our administrative assistant Ms Margery Knewstubb, the website developer Dr François DeBlois, the two heads of the logistics team, Ms Emily Poon and Ms Emily Heath, our local medical physics students and staff, the IOP staff and the authors who shared their new and exciting work with us.

Editors: Frank Verhaegen and Jan Seuntjens (McGill University) Associate editors: Luc Beaulieu, Iwan Kawrakow, Tony Popescu and David Rogers

Computed tomography (CT) images of patients with hip prostheses are severely degraded by metal streaking artefacts. The low image quality makes organ contouring more difficult and can result in large dose calculation errors when Monte Carlo (MC) techniques are used. In this work, the extent of streaking artefacts produced by three common hip prosthesis materials (Ti-alloy, stainless steel, and Co-Cr-Mo alloy) was studied. The prostheses were tested in a hypothetical prostate treatment with five 18 MV photon beams. The dose distributions for unilateral and bilateral prosthesis phantoms were calculated with the EGSnrc/DOSXYZnrc MC code. This was done in three phantom geometries: in the exact geometry, in the original CT geometry, and in an artefact-corrected geometry. The artefact-corrected geometry was created using a modified filtered back-projection correction technique. It was found that unilateral prosthesis phantoms do not show large dose calculation errors, as long as the beams miss the artefact-affected volume. This is possible to achieve in the case of unilateral prosthesis phantoms (except for the Co-Cr-Mo prosthesis which gives a 3% error) but not in the case of bilateral prosthesis phantoms. The largest dose discrepancies were obtained for the bilateral Co-Cr-Mo hip prosthesis phantom, up to 11% in some voxels within the prostate. The artefact correction algorithm worked well for all phantoms and resulted in dose calculation errors below 2%. In conclusion, a MC treatment plan should include an artefact correction algorithm when treating patients with hip prostheses.

Proton and carbon ion beams have a very sharp Bragg peak. For proton beams of energies smaller than 100 MeV, fitting with a gaussian the region of the maximum of the Bragg peak, the sigma along the beam direction is smaller than 1 mm, while for carbon ion beams, the sigma derived with the same technique is smaller than 1 mm for energies up to 360 MeV. In order to use low energy proton and carbon ion beams in hadrontherapy and to achieve an acceptable homogeneity of the spread out Bragg peak (SOBP) either the peak positions along the beam have to be quite close to each other or the longitudinal peak shape needs to be broaden at least few millimeters by means of a properly designed ripple filter. With a synchrotron accelerator in conjunction with active scanning techniques the use of a ripple filter is necessary to reduce the numbers of energy switches necessary to obtain a smooth SOBP, leading also to shorter overall irradiation times.

We studied the impact of the design of the ripple filter on the dose uniformity in the SOBP region by means of Monte Carlo simulations, implemented using the package Geant4. We simulated the beam delivery line supporting both proton and carbon ion beams using different energies of the beams. We compared the effect of different kind of ripple filters and their advantages.

A software graphical user interface (GUI) for calculation of 3D dose distribution using Monte Carlo (MC) simulation is developed using MATLAB. This GUI (DOSCTP) provides a user-friendly platform for DICOM CT-based dose calculation using EGSnrcMP-based DOSXYZnrc code. It offers numerous features not found in DOSXYZnrc, such as the ability to use multiple beams from different phase-space files, and has built-in dose analysis and visualization tools. DOSCTP is written completely in MATLAB, with integrated access to DOSXYZnrc and CTCREATE. The program function may be divided into four subgroups, namely, beam placement, MC simulation with DOSXYZnrc, dose visualization, and export. Each is controlled by separate routines. The verification of DOSCTP was carried out by comparing plans with different beam arrangements (multi-beam/photon arc) on an inhomogeneous phantom as well as patient CT between the GUI and Pinnacle³. DOSCTP was developed and verified with the following features: (1) a built-in voxel editor to modify CT-based DOSXYZnrc phantoms for research purposes; (2) multi-beam placement is possible, which cannot be achieved using the current DOSXYZnrc code; (3) the treatment plan, including the dose distributions, contours and image set can be exported to a commercial treatment planning system such as Pinnacle³ or to CERR using RTOG format for plan evaluation and comparison; (4) a built-in RTOG-compatible dose reviewer for dose visualization and analysis such as finding the volume of hot/cold spots in the 3D dose distributions based on a user threshold. DOSCTP greatly simplifies the use of DOSXYZnrc and CTCREATE, and offers numerous features that not found in the original user-code. Moreover, since phase-space beams can be defined and generated by the user, it is a particularly useful tool to carry out plans using specifically designed irradiators/accelerators that cannot be found in the Linac library of commercial treatment planning systems.

The Dose Planning Method (DPM) is one of several 'fast' Monte Carlo (MC) computer codes designed to produce an accurate dose calculation for advanced clinical applications. We have developed a flexible machine modeling process and validation tests for open-field and IMRT calculations. To complement the DPM code, a practical and versatile source model has been developed, whose parameters are derived from a standard set of planning system commissioning measurements. The primary photon spectrum and the spectrum resulting from the flattening filter are modeled by a Fatigue function, cut-off by a multiplying Fermi function, which effectively regularizes the difficult energy spectrum determination process. Commonly-used functions are applied to represent the off-axis softening, increasing primary fluence with increasing angle ('the horn effect'), and electron contamination. The patient dependent aspect of the MC dose calculation utilizes the multi-leaf collimator (MLC) leaf sequence file exported from the treatment planning system DICOM output, coupled with the source model, to derive the particle transport. This model has been commissioned for Varian 2100C 6 MV and 18 MV photon beams using percent depth dose, dose profiles, and output factors. A 3-D conformal plan and an IMRT plan delivered to an anthropomorphic thorax phantom were used to benchmark the model. The calculated results were compared to Pinnacle v7.6c results and measurements made using radiochromic film and thermoluminescent detectors (TLD).

Preclinical Microbeam Radiation Therapy (MRT) research programs are carried out at the European Synchrotron Radiation Facility (ESRF) and at a few other synchrotron facilities. MRT needs an accurate evaluation of the doses delivered to biological tissues for carrying out pre-clinical studies. This point is crucial for determining the effect induced by changing any of the physical irradiation parameters. The doses of interest in MRT are normally calculated using Monte Carlo (MC) methods. A few MC packages have been used in the last decade for MRT dose evaluations in independent studies. The aim of this investigation is to provide a preliminary basis to perform a systematic comparison of the dose results obtained, under identical irradiation conditions and for the same scoring geometries with the following five MC codes: EGS4, PENELOPE, GEANT4, EGSnrc, and MCNPX. Dose profiles have been calculated in an in-depth region of cylindrical phantoms made of water or PMMA. Beams in both cylindrical and planar geometry have been considered. This comparison shows an overall agreement among the different codes although minor differences occur, which need further investigations.

This work evaluates the effects of patient size on radiation dose from simulation imaging studies such as four-dimensional computed tomography (4DCT) and kilovoltage cone-beam computed tomography (kV-CBCT). 4DCT studies are scans that include temporal information, frequently incorporating highly over-sampled imaging series necessary for retrospective sorting as a function of respiratory phase. This type of imaging study can result in a significant dose increase to the patient due to the slower table speed as compared with a conventional axial or helical scan protocol. Kilovoltage cone-beam imaging is a relatively new imaging technique that requires an on-board kilovoltage x-ray tube and a flat-panel detector. Instead of porting individual reference fields, the kV tube and flat-panel detector are rotated about the patient producing a cone-beam CT data set (kV-CBCT). To perform these investigations, we used Monte Carlo simulation methods with detailed models of adult patients and virtual source models of multidetector computed tomography (MDCT) scanners. The GSF family of three-dimensional, voxelized patient models, were implemented as input files using the Monte Carlo code MCNPX. The adult patient models represent a range of patient sizes and have all radiosensitive organs previously identified and segmented. Simulated 4DCT scans of each voxelized patient model were performed using a multi-detector CT source model that includes scanner specific spectra, bow-tie filtration, and helical source path. Standard MCNPX tally functions were applied to each model to estimate absolute organ dose based upon an air-kerma normalization measurement for nominal scanner operating parameters.

Graphics processing units (GPUs) and similar stream processors are increasingly used for general-purpose calculations. Their pipelined architecture can be exploited to accelerate various algorithms, sometimes with spectacular results. Monte Carlo codes, being computationally intensive, are likely to benefit from the development of stream processing platforms. We explore this potential here with a simple subroutine sometimes used in Monte Carlo techniques. More specifically, a ray tracing algorithm that computes the exact radiological path in a voxel grid was implemented in CPU and GPU versions, which then were compared in terms of execution speed. This benchmarking experiment was conducted under various conditions, in order to assess the memory and bandwidth limitations of each platform. The results show that the GPU provides a significant speed improvement factor over the CPU. For the specific hardware used in this work, namely a nVidia 7600 GS GPU, a speed increase factor up to 6 was achieved over an Xeon 2.4 GHz CPU. With the development of faster stream processors, this factor is expected to reach levels that can potentially change how Monte Carlo techniques are used, for example in radiation therapy planning. The ongoing development of simpler language extensions and programming interfaces also promises to increase the accessibility of these devices. Overall, stream processors are likely to play an increasingly larger role in scientific computing, and in particular in Monte Carlo techniques.

In this study, a 4D treatment planning tool using an analytical model accounting for breathing motion is investigated to evaluate the motion effect on delivered dose for lung cancer treatments with three-dimensional conformal radiotherapy (3DCRT). The Monte Carlo EGS4/MCDOSE user code is used in the treatment planning dose calculation, and the patient CT data are converted into respective patient geometry files for Monte Carlo dose calculation. The model interpolates CT images at different phases of the breathing cycle from patient CT scans taken at end inspiration and end expiration phases and the chest wall position. Correlation between the voxels in a reference CT dataset and the voxels in the interpolated CT datasets at any breathing phases is established so that the dose to a voxel can be accumulated through the entire breathing cycle. Simulated lung tumors at different locations are used to demonstrate our model in 3DCRT for lung cancer treatments. We demonstrated the use of a 4D treatment planning tool in evaluating the breathing motion effect on delivered dose for different planning margins. Further studies are being conducted to use this tool to study the lung motion effect through large-scale analysis and to implement this useful tool for treatment planning dose calculation and plan evaluation for 4D radiotherapy.

The electron transport algorithm implemented in Geant4 has been recently revised. The modifications concern several physics aspects of the simulation model: the step limitation, the energy loss along a step and the multiple scattering. The Fano cavity setup was used to test these developments. The upgrades increase significantly the accuracy of the electron transport simulation. The ratio of simulated to theoretical dose deposition in the cavity is stable to ∼1% while varying several parameters and within ∼1.5% of the expected value for water and graphite. Work is underway to identify and resolve the remaining shift.

Monte Carlo (MC) calculations are rapidly finding their place in clinical dose assessments. We investigated conformal prostate dose distributions as calculated by MC, and compared them to several analytical dose calculations. The treatment distributions for twenty prostate cancer patients, treated with 18 MV 3D conformal radiation therapy, were retrospectively assessed. The BEAM code based on EGSnrc was used to model the beam from which phase space files were used as input into the XVMC algorithm. This was compared to conventional treatment planning system calculations (CADPLAN) with and without inhomogeneity corrections. Results indicate that the CADPLAN generalized Batho Power Law, modified Batho Power Law, and equivalent tissue-air ratio methods contain inaccuracies in calculated dose to 95 % of the prostate planning target volume of 3.5 %, 3.3 %, and 2.9 %, respectively. The greatest discrepancies in the organs at risk were seen in the bladder where the inhomogeneity correction methods all predicted that 50 % of the prescribed dose covered an average of 8.2 % more of the bladder volume than that predicted from the MC calculation. Water equivalent MC and water equivalent CADPLAN calculations revealed important discrepancies on the same order as those between heterogeneous MC and heterogeneous CADPLAN calculations. The data indicate that the effect of inhomogeneities is greater in the target volume than the organs at risk, and that accurately modeling the dose deposition process is important for each patient geometry, and may have a greater impact on the dose distribution in the prostate region than correcting an analytical algorithm for the presence of inhomogeneities.

At present, all clinical algorithms used in brachytherapy are based on the TG-43 algorithm, which has the advantage to offer very fast calculation time. However, this formalism has many simplifications, assuming for example the patient tissue composition equivalent to water. For low energy brachytherapy seeds such as iodine seeds, it is of interest to evaluate the dosimetric differences between calculations based on Monte Carlo simulations (considered the gold standard) and the TG-43 formalism. For a 6711 model ¹²⁵I seed calculated photon spectra were compared to spectra measured with a CdTe spectrometer. Good agreement was found except for the lowest energy peak which seems to be over-estimated by the experiment due to the contribution of the spectrometer CdTe diode to the measurement. Dose distributions in water are measured with EBT Gafchromic film and compared to the Monte Carlo calculation. A very good agreement is found. Finally, the method to create a MCNPX input file from computed tomography (CT) scanner images is explained and some preliminary isodose distributions are presented.

Brachytherapy is a radiotherapy treatment where encapsulated radioactive sources are introduced within a patient. Depending on the technique used, such sources can produce high, medium or low local dose rates. The Monte Carlo method is a powerful tool to simulate sources and devices in order to help physicists in treatment planning. In multiple types of gynaecological cancer, intracavitary brachytherapy (HDR Ir-192 source) is used combined with other therapy treatment to give an additional local dose to the tumour. Different types of applicators are used in order to increase the dose imparted to the tumour and to limit the effect on healthy surrounding tissues. The aim of this work is to model both applicator and HDR source in order to evaluate the dose at a reference point as well as the effect of the materials constituting the applicators on the near field dose. The MCNP5 code based on the Monte Carlo method has been used for the simulation. Dose calculations have been performed with ^*F8 energy deposition tally, taking into account photons and electrons. Results from simulation have been compared with experimental in-phantom dose measurements. Differences between calculations and measurements are lower than 5%.The importance of the source position has been underlined.

Dose calculation methods which incorporate tissue motion are an important tool for evaluating the effect of respiratory motion on the delivered dose distribution. 4D dose calculation methods use a sum of remapped doses calculated on 4D CT images of the patient at different respiratory phases to determine the cumulative dose received over the entire respiratory cycle. A number of methods for remapping the dose to the reference phase have been proposed, including center-of-mass (COM) tracking and trilinear (TL) interpolation. In this work we compare calculations of dose distributions remapped between extreme breathing phases against a 4D Monte Carlo dose code defDOSXYZ for three planning scenarios. No clinically significant differences were noted between dose distributions calculated by the three methods with the exception of an extreme motion evaluation case where TL and COM remapping underestimated the 95% target dose coverage by up to 16%. The accuracy of these dose calculation methods is significantly affected by the continuity of the deformation fields from non-linear image registration.

vmcPET, a VMC⁺⁺ based fast code for simulating photon transport through the patient geometry for use in positron emission tomography related calculations, is presented. vmcPET is shown to be between 250 and 425 times faster than GATE in completely analog mode and up to 50000 times faster when using advanced variance reduction techniques. Excellent agreement between vmcPET and EGSnrc and GATE benchmarks is found. vmcPET is coupled to GATE via phase-space files of particles emerging from the patient geometry.

The effects of high electron energy cutoff (ECUT) have been investigated and a method to reduce the dose statistical uncertainty caused by high ECUT was implemented in this work. In EGS4 Monte Carlo simulations, an electron is discarded and its energy is deposited locally when its total energy is lower than ECUT. The deposited energy can be significantly higher than the energy loss calculated using the CSDA model with the corresponding stopping powers in a low-density medium. This will create higher statistical uncertainties in the dose distributions, especially in air and lung tissues. In this work, a new method was implemented to continuously transport a discarded electron without considering multiple scattering or secondary particle generation. The energy loss is calculated based on the mass collision stopping powers in the local medium with an additional energy loss (70%) to account for the effect of approximations made in transporting the electron in a straight line rather than a curved path. The new method can significantly reduce the dose statistical uncertainty and thus improve the simulation efficiency even though the new method requires about 2% more CPU time. Our results showed that the statistical uncertainty of the dose in air cavities of a head-and-neck patient was reduced from 39% to about 2%. The dose distribution for the head and neck patient was significantly improved without losing dose accuracy and simulation efficiency.

Accurate dose calculation is essential to advanced stereotactic radiosurgery (SRS) and stereotactic radiotherapy (SRT) especially for treatment planning involving heterogeneous patient anatomy. This paper describes the implementation of a fast Monte Carlo dose calculation algorithm in SRS/SRT treatment planning for the CyberKnife® SRS/SRT system. A superposition Monte Carlo algorithm is developed for this application. Photon mean free paths and interaction types for different materials and energies as well as the tracks of secondary electrons are pre-simulated using the MCSIM system. Photon interaction forcing and splitting are applied to the source photons in the patient calculation and the pre-simulated electron tracks are repeated with proper corrections based on the tissue density and electron stopping powers. Electron energy is deposited along the tracks and accumulated in the simulation geometry. Scattered and bremsstrahlung photons are transported, after applying the Russian roulette technique, in the same way as the primary photons. Dose calculations are compared with full Monte Carlo simulations performed using EGS4/MCSIM and the CyberKnife treatment planning system (TPS) for lung, head & neck and liver treatments. Comparisons with full Monte Carlo simulations show excellent agreement (within 0.5%). More than 10% differences in the target dose are found between Monte Carlo simulations and the CyberKnife TPS for SRS/SRT lung treatment while negligible differences are shown in head and neck and liver for the cases investigated. The calculation time using our superposition Monte Carlo algorithm is reduced up to 62 times (46 times on average for 10 typical clinical cases) compared to full Monte Carlo simulations. SRS/SRT dose distributions calculated by simple dose algorithms may be significantly overestimated for small lung target volumes, which can be improved by accurate Monte Carlo dose calculations.

A very fast Monte Carlo algorithm for the calculation of the scatter contribution in cone beam computed tomography, implemented within the EGSnrc framework, is presented. Based on the combination of several variance reduction techniques, an efficiency improvement of three orders of magnitude over an analog simulation is obtained. A denoising algorithm applied to the computed scatter distribution is shown to further increase the efficiency of the calculation by about a factor of 10. An iterative scatter correction algorithm is proposed and its feasibility is demonstrated on three different phantoms.

A number of accelerated Monte Carlo (MC) codes have been developed in recent years for brachytherapy applications, one of which is PTRAN_CT. Developed as an extension to the well-benchmarked PTRAN code, PTRAN_CT can be used to perform efficient patient-specific dose calculations. The code can explicitly account for the patient geometry converted from computed-tomography (CT) images, as well as perturbations due to the brachytherapy applicator and seeds. We have developed a software tool called BrachyGUI that provides an integrated environment for preparing patient and treatment plan-specific input data files for PTRAN_CT. It also comes with dose calculation, analysis, and treatment planning capabilities. In this article, we will describe the interface of BrachyGUI with PTRAN_CT for CT-based calculations, and examine the calculation efficiency of PTRAN_CT. We conclude that it is now feasible to use PTRAN_CT for high dose rate brachytherapy treatment planning on a routine clinical basis.

Whilst a radiotherapy imaging system can be modeled accurately using full Monte Carlo simulations a quicker method is required for system optimisation. Here we present a Monte Carlo based optimisation model that takes a radiation spectrum and detector response curve as inputs and predicts contrast for a particular imaging system. The model consists of two parts. The first part looks at the interaction of mono-energetic beams with phantoms of various bone and water compositions. In-particular the scatter and primary components emerging from the phantom are analysed. The second part models the response of a detector to various mono-energetic beams. Weighting of the phantom simulations with a linac spectrum results in the prediction of the scatter and primary components incident on an imaging device. Subsequent application of a detector response curve yields the scatter and primary signals in the imager. This technique has been applied to standard 6 and 4MV linac spectra as well as an experimental low atomic number (Z) target configuration to determine various imaging parameters. In particular it has been used to determine the contrast with various detector, phantom and linac spectra combinations. Use of the model shows significant benefits in using a detector with a reduced metal plate thickness for the low Z beam for thin, 5.8cm phantoms that approximate the head and neck region. Whilst contrast can be doubled using the low Z beam and standard MV imager over the current 6MV system, a further improvement in contrast is predicted when using a detector without a metal plate.

Monte Carlo (MC) is rarely used for IMRT plan optimization outside of research centres due to the extensive computational resources or long computation times required to complete the process. Time can be reduced by degrading the statistical precision of the MC dose calculation used within the optimization loop. However, this eventually introduces optimization convergence errors (OCEs). This study determines the statistical noise levels tolerated during MC-IMRT optimization under the condition that the optimized plan has OCEs <100 cGy (1.5% of the prescription dose) for MC-optimized IMRT treatment plans.

Seven-field prostate IMRT treatment plans for 10 prostate patients are used in this study. Pre-optimization is performed for deliverable beams with a pencil-beam (PB) dose algorithm. Further deliverable-based optimization proceeds using: (1) MC-based optimization, where dose is recomputed with MC after each intensity update or (2) a once-corrected (OC) MC-hybrid optimization, where a MC dose computation defines beam-by-beam dose correction matrices that are used during a PB-based optimization. Optimizations are performed with nominal per beam MC statistical precisions of 2, 5, 8, 10, 15, and 20%. Following optimizer convergence, beams are re-computed with MC using 2% per beam nominal statistical precision and the 2 PTV and 10 OAR dose indices used in the optimization objective function are tallied. For both the MC-optimization and OC-optimization methods, statistical equivalence tests found that OCEs are less than 1.5% of the prescription dose for plans optimized with nominal statistical uncertainties of up to 10% per beam. The achieved statistical uncertainty in the patient for the 10% per beam simulations from the combination of the 7 beams is ∼3% with respect to maximum dose for voxels with D>0.5D_max. The MC dose computation time for the OC-optimization is only 6.2 minutes on a single 3 Ghz processor with results clinically equivalent to high precision MC computations.

The concave eye applicators with ¹⁰⁶Ru/¹⁰⁶Rh or ⁹⁰Sr/⁹⁰Y beta-ray sources are worldwide used in brachytherapy for treating intraocular tumors. It raises the need to know the exact dose delivered by beta radiation to tumors but measurement of the dose to water (or tissue) is very difficult due to short range of electrons. The Monte Carlo technique provides a powerful tool for calculation of the dose and dose distributions which helps to predict and determine the doses from different shapes of various types of eye applicators more accurately. The Monte Carlo code MCNPX has been used to calculate dose distributions from a COB-type ¹⁰⁶Ru/¹⁰⁶Rh ophthalmic applicator manufactured by Eckert & Ziegler BEBIG GmbH. This type of a concave eye applicator has a cut-out whose purpose is to protect the eye nerve which makes the dose distribution more complicated. Several calculations have been performed including depth dose along the applicator central axis and various dose distributions. The depth dose along the applicator central axis and the dose distribution on a spherical surface 1 mm above the plaque inner surface have been compared with measurement data provided by the manufacturer. For distances from 0.5 to 4 mm above the surface, the agreement was within 2.5 % and from 5 mm the difference increased from 6 % up to 25 % at 10 mm whereas the uncertainty on manufacturer data is 20 % (2s). It is assumed that the difference is caused by nonuniformly distributed radioactivity over the applicator radioactive layer.

Helical Tomotherapy (HT) delivers intensity-modulated radiotherapy by the means of many configurations of the binary multi-leaf collimator (MLC). The aim of the present study was to devise a method, which we call the 'transfer function' (TF) method, to perform the transport of particles through the MLC much faster than the time consuming Monte Carlo (MC) simulation and with no significant loss of accuracy. The TF method consists of calculating, for each photon in the phase-space file, the attenuation factor for each leaf (up to three) that the photon passes, assuming straight propagation through closed leaves, and storing these factors in a modified phase-space file. To account for the transport through the MLC in a given configuration, the weight of a photon is simply multiplied by the attenuation factors of the leaves that are intersected by the photon ray and are closed. The TF method was combined with the PENELOPE MC code, and validated with measurements for the three static field sizes available (40×5, 40×2.5 and 40×1 cm²) and for some MLC patterns. The TF method allows a large reduction in computation time, without introducing appreciable deviations from the result of full MC simulations.

Novel technologies aiming at improving target dose coverage while minimising dose to organs at risk use delivery of radiation fields that significantly deviate from reference conditions defined in protocols such as TG-51 and TRS-398. The use of ionization chambers for patient-specific quality assurance of these new delivery procedures calibrated in reference conditions increases the uncertainties on dose delivery. The conversion of the dose to the chamber cavity to the dose to water becomes uncertain; and the geometrical details of the chamber, as well as the details of the delivery, are expected to be significant. In this study, a realistic model of the Exradin® A12 Farmer chamber is simulated. A framework is applied for the calculation of ionization chamber response to arbitrarily modulated fields as a summation of responses to pencil beams. This approach is used with the chamber model and tested against measurements in static open fields and dynamic MLC IMRT fields. As a benchmark test of the model, quality conversion factors values calculated by Monte-Carlo simulation with the chamber model are in agreement within 0.1 % and 0.4 % with those in the AAPM TG-51, for 6 MV and 18 MV photon beams, respectively. Pencil-beam kernels show a strong dependence on the geometrical details of the chamber. Kernel summations with open fields show a relative agreement within 4.0 % with experimental data; the agreement is within 2.0 % for dynamic MLC IMRT beams. Simulations show a strong sensitivity of chamber response on positioning uncertainties, sometimes leading to dose uncertainties of 15 %.

The use of Monte Carlo methods in photon beam treatment planning is becoming feasible due to advances in hardware and algorithms. However, a major challenge is the modeling of the radiation produced by individual linear accelerators. Monte Carlo simulation through the accelerator head or a parameterized source model may be used for this purpose. In this work, the latter approach was chosen due to larger flexibility and smaller amount of required information about the accelerator composition. The source model used includes sub-sources for primary photons emerging from target, extra-focal photons, and electron contamination. The free model parameters were derived by minimizing an objective function measuring deviations between pencil-beam-kernel based dose calculations and measurements. The output of the source model was then used as input for the VMC++ code, which was used to transport the particles through the accessory modules and the patient. To verify the procedure, VMC++ calculations were compared to measurements for open, wedged, and irregular MLC-shaped fields for 6MV and 15MV beams. The observed discrepancies were mostly within 2%, 2 mm. This work demonstrates that the developed procedure could, in the future, be used to commission the VMC++ algorithm for clinical use in a hospital.

IMRT often requires delivering small fields which may suffer from electronic disequilibrium effects. The presence of heterogeneities, particularly low-density tissues in patients, complicates such situations. In this study, we report on verification of the DPM MC code for IMRT treatment planning in heterogeneous media, using a previously developed model of the Varian 120-leaf MLC. The purpose of this study is twofold: (a) design a comprehensive list of experiments in heterogeneous media for verification of any dose calculation algorithm and (b) verify our MLC model in these heterogeneous type geometries that mimic an actual patient geometry for IMRT treatment. The measurements have been done using an IMRT head and neck phantom (CIRS phantom) and slab phantom geometries. Verification of the MLC model has been carried out using point doses measured with an A14 slim line (SL) ion chamber inside a tissue-equivalent and a bone-equivalent material using the CIRS phantom. Planar doses using lung and bone equivalent slabs have been measured and compared using EDR films (Kodak, Rochester, NY).

Diagnostic imaging with CT procedures is responsible for significant radiation doses to patients. To enable individual patient dose estimates, a combination of MOSFET detectors and Monte Carlo (MC) simulations was investigated for the determination of patient surface dose. The behaviour of MOSFETs in kV x-rays from a CT scanner was investigated with experiments and MC simulations with a CT scanner model. A dose reproducibility of 5% and a mean loss of sensitivity with accumulated dose of about 10% was noted for the MOSFETs. Beam energy increase from 80-140 kVp resulted in a response decrease of 10%. The MOSFET detectors were calibrated in terms of absolute surface dose with the aid of MC simulations. Good agreement was achieved between measured and calculated surface dose on a cylindrical Lucite phantom. Experiments with a stationary x-ray tube and contiguous axial scanning led to differences limited by 8%. Surface dose in helical scanning was investigated with measurements with radiological film and an array of five MOSFET detectors, leading to good agreement. It is concluded that an array of MOSFET detectors, calibrated in terms of surface dose, is a valuable tool to assess individual patient surface dose. In combination with MC simulations this may lead to estimations of effective dose.

Modern treatment planning systems (TPSs) usually separate the dose modelling into a beam modelling phase, describing the beam exiting the accelerator, followed by a subsequent dose calculation in the patient. The aim of this work is to use the Monte Carlo code system EGSnrc to study the modelling of head scatter as well as the transmission through multi-leaf collimator (MLC) and diaphragms in the beam model used in a commercial TPS (MasterPlan, Nucletron B.V.). An Elekta Precise linear accelerator equipped with an MLC has been modelled in BEAMnrc, based on available information from the vendor regarding the material and geometry of the treatment head. The collimation in the MLC direction consists of leafs which are complemented with a backup diaphragm. The characteristics of the electron beam, i.e., energy and spot size, impinging on the target have been tuned to match measured data. Phase spaces from simulations of the treatment head are used to extract the scatter from, e.g., the flattening filter and the collimating structures. Similar data for the source models used in the TPS are extracted from the treatment planning system, thus a comprehensive analysis is possible. Simulations in a water phantom, with DOSXYZnrc, are also used to study the modelling of the MLC and the diaphragms by the TPS. The results from this study will be helpful to understand the limitations of the model in the TPS and provide knowledge for further improvements of the TPS source modelling.