Experimental assessment of proton dose calculation accuracy in inhomogeneous media

PurposeProton therapy with Pencil Beam Scanning (PBS) has the potential to improve radiotherapy treatments. Unfortunately, its promises are jeopardized by the sensitivity of the dose distributions to uncertainties, including dose calculation accuracy in inhomogeneous media. Monte Carlo dose engines (MC) are expected to handle heterogeneities better than analytical algorithms like the pencil-beam convolution algorithm (PBA). In this study, an experimental phantom has been devised to maximize the effect of heterogeneities and to quantify the capability of several dose engines (MC and PBA) to handle these.MethodsAn inhomogeneous phantom made of water surrounding a long insert of bone tissue substitute (1 × 10 × 10 cm³) was irradiated with a mono-energetic PBS field (10 × 10 cm²). A 2D ion chamber array (MatriXX, IBA Dosimetry GmbH) lied right behind the bone. The beam energy was such that the expected range of the protons exceeded the detector position in water and did not attain it in bone. The measurement was compared to the following engines: Geant4.9.5, PENH, MCsquare, as well as the MC and PBA algorithms of RayStation (RaySearch Laboratories AB).ResultsFor a γ-index criteria of 2%/2 mm, the passing rates are 93.8% for Geant4.9.5, 97.4% for PENH, 93.4% for MCsquare, 95.9% for RayStation MC, and 44.7% for PBA. The differences in γ-index passing rates between MC and RayStation PBA calculations can exceed 50%.ConclusionThe performance of dose calculation algorithms in highly inhomogeneous media was evaluated in a dedicated experiment. MC dose engines performed overall satisfactorily while large deviations were observed with PBA as expected.     

Comparison of the RayStation photon Monte Carlo dose calculation algorithm against measured data under homogeneous and heterogeneous irradiation geometries

PurposeThis work compares Monte Carlo dose calculations performed using the RayStation treatment planning system against data measured on a Varian Truebeam linear accelerator with 6 MV and 10 MV FFF photon beams.MethodsThe dosimetric performance of the RayStation Monte Carlo calculations was evaluated in a variety of irradiation geometries employing homogeneous and heterogeneous phantoms. Profile and depth dose comparisons against measurement were carried out in relative mode using the gamma index as a quantitative measure of similarity within the central high dose regions.ResultsThe results demonstrate that the treatment planning system dose calculation engine agrees with measurement to within 2%/1 mm for more than 95% of the data points in the high dose regions for all test cases. A systematic underestimation was observed at the tail of the profile penumbra and out of field, with mean differences generally <0.5 mm or 1% of curve dose maximum respectively. Out of field agreement varied between evaluated beam models.ConclusionsThe RayStation implementation of photon Monte Carlo dose calculations show good agreement with measured data for the range of scenarios considered in this work and is deemed sufficiently accurate for introduction into clinical use.     

Accurate proton treatment planning for pencil beam crossing titanium fixation implants

PurposeTo present a planning strategy for proton pencil-beam scanning when titanium implants need to be crossed by the beam.MethodsWe addressed three issues: the implementation of a CT calibration curve to assign to titanium the correct stopping power; the effect of artefacts on CT images and their reduction by a dedicated algorithm; the differences in dose computation depending on the dose engine, pencil-beam vs Monte-Carlo algorithms. We performed measurement tests on a simple cylinder phantom and on a real implant. These phantoms were irradiated with three geometries (single spots, uniform mono-energetic layer and uniform box), measuring the exit dose either by radio-chromic film or multi-layer ionization chamber. The procedure was then applied on two patients treated for chordoma.ResultsWe had to set in the calibration curve a mass density equal to 4.37 g/cm³ to saturated Hounsfield Units, in order to have the correct stopping power assigned to titanium in TPS. CT artefact reduction algorithm allowed a better reconstruction of the shape and size of the implant. Monte-Carlo resulted accurate in computing the dose distribution whereas the pencil-beam algorithm failed due to sharp density interfaces between titanium and the surrounding material. Finally, the treatment plans obtained on two patients showed the impact of the dose engine algorithm, with 10–20% differences between pencil-beam and Monte-Carlo in small regions distally to the titanium screws.ConclusionThe described combination of CT calibration, artefacts reduction and Monte-Carlo computation provides a reliable methodology to compute dose in patients with titanium implants.     

FLUKA simulation of target fragmentation in proton therapy

In proton therapy, secondary fragments are created in nuclear interactions of the beam with the target nuclei. The secondary fragments have low kinetic energies and high atomic numbers as compared to primary protons. Fragments have a high LET and deposit all their energy close to the generation point. For their characteristics, secondary fragments can alter the dose distribution and lead to an increase of RBE for the same delivered physical dose. Moreover, the radiobiological impact of target fragmentation is significant mostly in the region before the Bragg peak, where generally healthy tissues are present, and immediately after Bragg peak. Considering the high biological impact of those particles, especially in the case of healthy tissues or organs at risk, the inclusion of target fragmentation processes in the dose calculation of a treatment planning system can be relevant to improve the treatment accuracy and for this reason it is one of the major tasks of the MoVe IT project.In this study, Monte Carlo simulations were employed to fully characterize the mixed radiation field generated by target fragmentation in proton therapy. The dose averaged LET has been evaluated in case of a Spread Out Bragg Peak (SOBP). Starting from LET distribution, RBE has been evaluated with two different phenomenological models. In order to characterize the mixed radiation field, the production cross section has been evaluated by means of the FLUKA code. The future development of present work is to generate a MC database of fragments fluence to be included in TPS.     

Local dose enhancement of proton therapy by ceramic oxide nanoparticles investigated with Geant4 simulations

Nanoparticles (NPs) have been shown to enhance X-ray radiotherapy and proton therapy of cancer. The effectiveness of radiation damage is enhanced in the presence of high atomic number (high-Z) NPs due to increased production of low energy, higher linear energy transfer (LET) secondary electrons when NPs are selectively internalized by tumour cells. This work quantifies the local dose enhancement produced by the high-Z ceramic oxide NPs Ta₂O₅ and CeO₂, in the target tumour, for the first time in proton therapy, by means of Geant4 simulations. The dose enhancement produced by the ceramic oxides is compared against gold NPs. The energy deposition on a nanoscale around a single nanoparticle of 100 nm diameter is investigated using the Geant4-DNA extension to model particle interactions in the water medium. Enhancement of energy deposition in nano-sized shells of water, local to the NP boundary, ranging between 14% and 27% was observed for proton energies of 5 MeV and 50 MeV, depending on the NP material. Enhancement of electron production and energy deposition can be correlated to the direct DNA damage mechanism if the NP is in close proximity to the nucleus.     

Measurement of stray neutron doses inside the treatment room from a proton pencil beam scanning system

PurposeTo measure the environmental doses from stray neutrons in the vicinity of a solid slab phantom as a function of beam energy, field size and modulation width, using the proton pencil beam scanning (PBS) technique.MethodMeasurements were carried out using two extended range WENDI-II rem-counters and three tissue equivalent proportional counters. Detectors were suitably placed at different distances around the RW3 slab phantom. Beam irradiation parameters were varied to cover the clinical ranges of proton beam energies (100–220 MeV), field sizes ((2 × 2)–(20 × 20) cm²) and modulation widths (0–15 cm).ResultsFor pristine proton peak irradiations, large variations of neutron H¹(10)/D were observed with changes in beam energy and field size, while these were less dependent on modulation widths. H¹(10)/D for pristine proton pencil beams varied between 0.04 μSv Gy⁻¹ at beam energy 100 MeV and a (2 × 2) cm² field at 2.25 m distance and 90° angle with respect to the beam axis, and 72.3 μSv Gy⁻¹ at beam energy 200 MeV and a (20 × 20) cm² field at 1 m distance along the beam axis.ConclusionsThe obtained results will be useful in benchmarking Monte Carlo calculations of proton radiotherapy in PBS mode and in estimating the exposure to stray radiation of the patient. Such estimates may be facilitated by the obtained best-fitted simple analytical formulae relating the stray neutron doses at points of interest with beam irradiation parameters.     

The determination of a dose deposited in reference medium due to (p,n) reaction occurring during proton therapy

AimThe aim of the investigation was to determine the undesirable dose coming from neutrons produced in reactions (p,n) in irradiated tissues represented by water.BackgroundProduction of neutrons in the system of beam collimators and in irradiated tissues is the undesirable phenomenon related to the application of protons in radiotherapy. It makes that proton beams are contaminated by neutrons and patients receive the undesirable neutron dose.Materials and methodsThe investigation was based on the Monte Carlo simulations (GEANT4 code). The calculations were performed for five energies of protons: 50 MeV, 55 MeV, 60 MeV, 65 MeV and 75 MeV. The neutron doses were calculated on the basis of the neutron fluence and neutron energy spectra derived from simulations and by means of the neutron fluence–dose conversion coefficients taken from the ICRP dosimetry protocol no. 74 for the antero-posterior irradiation geometry.ResultsThe obtained neutron doses are much less than the proton ones. They do not exceed 0.1%, 0.4%, 0.5%, 0.6% and 0.7% of the total dose at a given depth for the primary protons with energy of 50 MeV, 55 MeV, 60 MeV, 65 MeV and 70 MeV, respectively.ConclusionsThe neutron production takes place mainly along the central axis of the beam. The maximum neutron dose appears at about a half of the depth of the maximum proton dose (Bragg peak), i.e. in the volume of a healthy tissue. The doses of neutrons produced in the irradiated medium (water) are about two orders of magnitude less than the proton doses for the considered range of energy of protons.     

Clinical implications of different calculation algorithms in breast radiotherapy: A comparison between pencil beam and collapsed cone convolution

BackgroundThis investigation focused on the clinical implications of the use of the Collapsed Cone Convolution algorithm (CCC) in breast radiotherapy and investigated the dosimetric differences as respect to Pencil Beam Convolution algorithm (PBC).Material and methods15 breast treatment plans produced using the PBC algorithm were re-calculated using the CCC algorithm with the same MUs. In a second step, plans were re-optimized using CCC algorithm with modification of wedges and beam weightings to achieve optimal coverage (CCCr plans). For each patient, dosimetric comparison was performed using the standard tangential technique (SWT) and a forward-planned IMRT technique (f-IMRT).ResultsThe CCC algorithm showed significant increased dose inhomogeneity. Mean and minimum PTV doses decreased by 1.4% and 2.8% (both techniques). Mean V95% decreased to 83.7% and 90.3%, respectively for the SWT and f-IMRT. V95% was correlated to the ratio of PTV and lung volumes into the treatment field. The re-optimized CCCr plans achieved similar target coverage, but high-dose volume was significantly larger (V107%: 7.6% vs 2.3% (SWT), 7.1% vs 2.1% (f-IMRT). There was a significantly increase in the ipsilateral lung volume receiving low doses (V5 Gy: 31.3% vs 26.2% in SWT, 27.0% vs 23.0% in f-IMRT). MUs needed for PTV coverage in CCCr plans were higher by 3%.ConclusionsThe PBC algorithm overestimated PTV coverage in terms of all important dosimetric metrics. If previous clinical experience are based on the use of PBC model, especially needed is discussion between medical physicists and radiation oncologists to fully understand the dosimetric changes.     

Out-of-field doses from secondary radiation produced in proton therapy and the associated risk of radiation-induced cancer from a brain tumor treatment

PurposeTo determine out-of-field doses produced in proton pencil beam scanning (PBS) therapy using Monte Carlo simulations and to estimate the associated risk of radiation-induced second cancer from a brain tumor treatment.MethodsSimulations of out-of-field absorbed doses were performed with MCNP6 and benchmarked against measurements with tissue-equivalent proportional counters (TEPC) for three irradiation setups: two irradiations of a water phantom using proton energies of 78–147 MeV and 177–223 MeV, and one brain tumor irradiation of a whole-body phantom. Out-of-field absorbed and equivalent doses to organs in a whole-body phantom following a brain tumor treatment were subsequently simulated and used to estimate the risk of radiation-induced cancer. Additionally, the contribution of absorbed dose originating from radiation produced in the nozzle was calculated from simulations.ResultsOut-of-field absorbed doses to the TEPC ranged from 0.4 to 135 µGy/Gy. The average deviation between simulations and measurements of the water phantom irradiations was about 17%. The absorbed dose contribution from radiation produced in the nozzle ranged between 0 and 70% of the total dose; the contribution was however small in absolute terms. The absorbed and equivalent doses to the organs ranged between 0.2 and 60 µGy/Gy and 0.5–151 µSv/Gy. The estimated lifetime risk of radiation-induced second cancer was approximately 0.01%.ConclusionsThe agreement of out-of-field absorbed doses between measurements and simulations was good given the sources of uncertainties. Calculations of out-of-field organ doses following a brain tumor treatment indicated that proton PBS therapy of brain tumors is associated with a low risk of radiation-induced cancer.     

Configuration and validation of an analytical model predicting secondary neutron radiation in proton therapy using Monte Carlo simulations and experimental measurements

PurposeThis study focuses on the configuration and validation of an analytical model predicting leakage neutron doses in proton therapy.MethodsUsing Monte Carlo (MC) calculations, a facility-specific analytical model was built to reproduce out-of-field neutron doses while separately accounting for the contribution of intra-nuclear cascade, evaporation, epithermal and thermal neutrons. This model was first trained to reproduce in-water neutron absorbed doses and in-air neutron ambient dose equivalents, H*(10), calculated using MCNPX. Its capacity in predicting out-of-field doses at any position not involved in the training phase was also checked. The model was next expanded to enable a full 3D mapping of H*(10) inside the treatment room, tested in a clinically relevant configuration and finally consolidated with experimental measurements.ResultsFollowing the literature approach, the work first proved that it is possible to build a facility-specific analytical model that efficiently reproduces in-water neutron doses and in-air H*(10) values with a maximum difference less than 25%. In addition, the analytical model succeeded in predicting out-of-field neutron doses in the lateral and vertical direction. Testing the analytical model in clinical configurations proved the need to separate the contribution of internal and external neutrons. The impact of modulation width on stray neutrons was found to be easily adjustable while beam collimation remains a challenging issue. Finally, the model performance agreed with experimental measurements with satisfactory results considering measurement and simulation uncertainties.ConclusionAnalytical models represent a promising solution that substitutes for time-consuming MC calculations when assessing doses to healthy organs.     

Simulation of dose distribution and secondary particle production in proton therapy of brain tumor

AimThe aim of this study is simulation of the proton depth-dose distribution and dose evaluation of secondary particles in proton therapy of brain tumor using the GEANT4 and FLUKA Monte Carlo codes.BackgroundProton therapy is a treatment method for variety of tumors such as brain tumor. The most important feature of high energy proton beams is the energy deposition as a Bragg curve and the possibility of creating the spread out Bragg peak (SOBP) for full coverage of the tumor.Materials and methodsA spherical tumor with the radius of 1 cm in the brain is considered. A SNYDER head phantom has been irradiated with 30−130 MeV proton beam energy. A PMMA modulator wheel is used for covering the tumor. The simulations are performed using the GEANT4 and FLUKA codes.ResultsUsing a modulator wheel, the Spread Out Bragg Peak longitudinally and laterally covers the tumor. Flux and absorbed dose of secondary particles produced by nuclear interactions of protons with elements in the head are considerably small compared to protons.ConclusionsUsing 76.85 MeV proton beam and a modulator wheel, the tumor can be treated accurately in the 3-D, so that the distribution of proton dose in the surrounding tissues is very low. The results show that more than 99% of the total dose of secondary particles and protons is absorbed in the tumor.     

Beam characterisation studies of the 62 MeV proton therapy beamline at the Clatterbridge Cancer Centre

The Clatterbridge Cancer Centre (CCC) in the United Kingdom is the world’s first hospital proton beam therapy facility, providing treatment for ocular cancers since 1989. A 62 MeV beam of protons is produced by a Scanditronix cyclotron and transported through a passive delivery system. In addition to the long history of clinical use, the facility supports a wide programme of experimental work and as such, an accurate and reliable simulation model of the treatment beamline is highly valuable. However, as the facility has seen several changes to the accelerator and beamline over the years, a comprehensive study of the CCC beam dynamics is needed to firstly examine the beam optics. An extensive analysis was required to overcome facility related constraints to determine fundamental beamline parameters and define an optical lattice written with the Methodical Accelerator Design (MAD-X) and the particle tracking Beam Delivery Simulation (BDSIM) code. An optimised case is presented and simulated results of the optical functions, beam distribution, losses and the transverse rms beam sizes along the beamline are discussed. Corresponding optical and beam information was used in TOPAS to simulate transverse beam profiles and compared to EBT3 film measurements. We provide an overview of the magnetic components, beam transport, cyclotron, beam and treatment related parameters necessary for the development of a present day optical model of the facility. This work represents the first comprehensive study of the CCC facility to date, as a basis to determine input beam parameters to accurately simulate and completely characterise the beamline.     

Characterization of a commercial scintillation detector for 2-D dosimetry in scanned proton and carbon ion beams

IntroductionPencil beam scanning technique used at CNAO requires beam characteristics to be carefully assessed and periodically checked to guarantee patient safety. This study aimed at characterizing the Lynx® detector (IBA Dosimetry) for commissioning and periodic quality assurance (QA) for proton and carbon ion beams, as compared to EBT3 films, currently used for QA checks.Methods and materialsThe Lynx® is a 2-D high-resolution dosimetry system consisting of a scintillating screen coupled with a CCD camera, in a compact light-tight box. The scintillator was preliminarily characterized in terms of short-term stability, linearity with number of particles, image quality and response dependence on iris setting and beam current; Lynx® was then systematically tested against EBT3 films. The detector response dependence on radiation LET was also assessed.ResultsPreliminary results have shown that Lynx is suitable to be used for commissioning and QA checks for proton and carbon ion scanning beams; the cross-check with EBT3 films showed a good agreement between the two detectors, for both single spot and scanned field measurements. The strong LET dependence of the scintillator due to quenching effect makes Lynx® suitable only for relative 2-D dosimetry measurements.ConclusionLynx® appears as a promising tool for commissioning and periodic QA checks for both protons and carbon ion beams. This detector can be used as an alternative of EBT3 films, allowing real-time measurements and analysis, with a significant time sparing.     

On the parametrization of lateral dose profiles in proton radiation therapy

PurposeThe accurate evaluation of the lateral dose profile is an important issue in the field of proton radiation therapy. The beam spread, due to Multiple Coulomb Scattering (MCS), is described by the Molière's theory. To take into account also the contribution of nuclear interactions, modern Treatment Planning Systems (TPSs) generally approximate the dose profiles by a sum of Gaussian functions. In this paper we have compared different parametrizations for the lateral dose profile of protons in water for therapeutical energies: the goal is to improve the performances of the actual treatment planning.MethodsWe have simulated typical dose profiles at the CNAO (Centro Nazionale di Adroterapia Oncologica) beamline with the FLUKA code and validated them with data taken at CNAO considering different energies and depths. We then performed best fits of the lateral dose profiles for different functions using ROOT and MINUIT.ResultsThe accuracy of the best fits was analyzed by evaluating the reduced χ², the number of free parameters of the functions and the calculation time. The best results were obtained with the triple Gaussian and double Gaussian Lorentz–Cauchy functions which have 6 parameters, but good results were also obtained with the so called Gauss–Rutherford function which has only 4 parameters.ConclusionsThe comparison of the studied functions with accurate and validated Monte Carlo calculations and with experimental data from CNAO lead us to propose an original parametrization, the Gauss–Rutherford function, to describe the lateral dose profiles of proton beams.     

A depth dose study between AAA and AXB algorithm against Monte Carlo simulation using AIP CT of a 4D dataset from a moving phantom

Aim

To identifying depth dose differences between the two versions of the algorithms using AIP CT of a 4D dataset.

A semi-analytical model for calculating the penetration depth of a high energy electron beam in a water phantom with a magnetic field

PurposeAs an electron beam is incident on a uniform water phantom in the presence of a lateral magnetic field, the depth-dose distribution of the electron beam changes significantly and forms the well-known ‘Bragg peak’, with a depth-dose distribution similar to that of heavy ions. This phenomenon has pioneered a new field in the clinical application of electron beams. For such clinical applications, evaluating the penetration depth of electron beams quickly and accurately is the critical problem.MethodsThis paper describes a model for calculating the penetration depth of an electron beam rapidly and correctly in a water phantom under the influence of a magnetic field. The model was used to calculate the penetration depths under different conditions: the energies of electron beams of 6, 8, 12 and 15 MeV and the magnetic induction intensities of 0.75, 1.0, 1.5, 2.0 and 3.0 T. In addition, the calculation results were compared with the results of a Monte Carlo simulation.ResultsThe comparison results indicate that the difference between the two calculation methods was less than 0.5 cm. Moreover, the computing time of the calculation model was less than a second.ConclusionsThe semi-analytical model proposed in the present study enables the penetration depth of the electron beam in the presence of a magnetic field to be obtained with a computational efficiency higher than that of the Monte Carlo approach; thus, the proposed model has high potential for application.     

A review of dose calculation approaches with cone beam CT in photon and proton therapy

Background and purposeThe use of cone beam computed tomography (CBCT) for performing dose calculations in radiation therapy has been widely investigated as it could provide a quantitative analysis of the dosimetric impact of changes in patients during the treatment. The aim of this review was to classify different techniques adopted to perform CBCT dose calculation and to report their dosimetric accuracy with respect to the metrics used.Methods and materialsA literature search was carried out in PubMed and ScienceDirect databases, based upon the following keywords: “cone beam computed tomography”, “CBCT”, “cone beam CT”, “dose calculation”, “accuracy”. Sixty-nine peer-reviewed relevant articles were included in this review: thirty-one patient studies, fifteen phantom studies and twenty-three patient & phantom studies. Most studies were found to have focused on head and neck, lung and prostate cancers.ResultsThe techniques adopted to perform CBCT dose calculation have been grouped in six categories labelled as (1) pCT calibration, (2) CBCT calibration, (3) HU override, (4) Deformable image registration, (5) Dose deformation, and (6) Combined techniques. Differences between CBCT dose and reference dose were reported both for target volumes and OARs.ConclusionsA comparison among the available techniques for CBCT dose calculations is challenging as many variables are involved. Therefore, a set of reporting standards is recommended to enable meaningful comparisons among different studies. The accuracy of the results was strongly dependent on the image quality, regardless of the methods used, highlighting the need for dose validation and quality assurance standards.