Extended in-field and out-of-field validation of a compact Monte Carlo model of an IBA PROTEUS®ONE proton beam in TOPAS/GEANT4.

Objective:This study evaluates a compact Monte Carlo (MC) model of a pencil beam scanning clinical proton beam using TOPAS to estimate the dose out-of-field (OOF). Compact modelling means that the model starts from a pristine proton beam at the nozzle exit, customised based on acceptance and commissioning data, instead of modelling the full treatment head and room.Approach: First, in-field validation tests were performed. Then, the OOF dose was validated in an RW3 phantom with bubble detectors for personal neutron dosimetry (measuring the neutron dose equivalent) and thermoluminiescent detectors (measuring the absorbed dose by protons and gammas). Measurements were performed at 15 and 35 cm from the distal edge of the field for five different irradiation plans, covering different beam orientations, proton energies and a 40 mm range shifter. TOPAS simulations were performed with QGSP Binary Cascade HP (BIC) and QGSP Bertini HP (Bertini) hadron physics lists.Main results: In-field validation shows that MC simulations agree with point dose measurements within -2.5 % and +1.5 % at locations on- and off-axis and before, in and after the Bragg peak or plateau. The gamma passing rate 2%/3mm of four simulated treatment plans compared to the dose distribution calculated by the TPS exceeds 97 % agreement score. OOF dose simulations showed an average overestimation of 27 % of the neutron dose equivalent for the BIC hadron physics list and an average underestimation of 20 % for the Bertini hadron physics list. The simulated absorbed dose of protons and gammas showed a systematic underestimation which was on average 21 % and 51 % for BIC and Bertini respectively.Significance: Our study demonstrates that a compact MC model can reliably produce in-field data, while out-of-field dose data are within the uncertainties of the detector systems and MC simulations nuclear models, and do so with shorter modelling and faster calculation time.

TOPAS simulations of the response of a mini-TEPC: benchmark with experimental data.

Objective. Microdosimetry offers a fast tool for radiation quality (RQ) verification to be implemented in treatment planning systems in proton therapy based on variable LET or RBE to move forward from the use of a fixed RBE of 1.1. It is known that the RBE of protons can increase up to 50% higher than that value in the last few millimetres of their range. Microdosimetry can be performed both experimentally and by means of Monte Carlo (MC) simulations. This paper has the aim of comparing the two approaches.Approach. Experimental measurements have been performed using a miniaturized Tissue equivalent proportional counter developed at the Legnaro National Laboratories of the Italian National Institute for Nuclear Physics with the aim of being used as RQ monitors for high intensity beams. MC simulations have been performed using the microdosimetric extension of TOPAS which provides optimized parameters and scorers for this application.Main results. Simulations were compared with experimental microdosimetric spectra in terms of shape of the spectra and their average values. Moreover, the latter have been investigated as possible estimators of LET obtained with the same MC code. The shape of the spectra is in general consistent with the experimental distributions and the average values of the distributions in both cases can predict the RQ increase with depth.Significance. This study aims at the comparison of microdosimetric spectra obtained from both experimental measurements and the microdosimetric extension of TOPAS in the same radiation field.

A deep learning and Monte Carlo based framework for bioluminescence imaging center of mass-guided glioblastoma targeting.

Objective.Bioluminescence imaging (BLI) is a valuable tool for non-invasive monitoring of glioblastoma multiforme (GBM) tumor-bearing small animals without incurring x-ray radiation burden. However, the use of this imaging modality is limited due to photon scattering and lack of spatial information. Attempts at reconstructing bioluminescence tomography (BLT) using mathematical models of light propagation show limited progress.Approach.This paper employed a different approach by using a deep convolutional neural network (CNN) to predict the tumor's center of mass (CoM). Transfer-learning with a sizeable artificial database is employed to facilitate the training process for, the much smaller, target database including Monte Carlo (MC) simulations of real orthotopic glioblastoma models. Predicted CoM was then used to estimate a BLI-based planning target volume (bPTV), by using the CoM as the center of a sphere, encompassing the tumor. The volume of the encompassing target sphere was estimated based on the total number of photons reaching the skin surface.Main results.Results show sub-millimeter accuracy for CoM prediction with a median error of 0.59 mm. The proposed method also provides promising performance for BLI-based tumor targeting with on average 94% of the tumor inside the bPTV while keeping the average healthy tissue coverage below 10%.Significance.This work introduced a framework for developing and using a CNN for targeted radiation studies for GBM based on BLI. The framework will enable biologists to use BLI as their main image-guidance tool to target GBM tumors in rat models, avoiding delivery of high x-ray imaging dose to the animals.

Advanced design, simulation, and dosimetry of a novel rectal applicator for contact brachytherapy with a conventional HDR 192Ir source.

PURPOSE: Dose escalation yields higher complete response to rectal tumors, which may enable the omission of surgery. Dose escalation using 50 kVp contact x-ray brachytherapy (CXB) allow the treatment of a selective volume, resulting in low toxicity and organs-at-risk preservation. However, the use of CXB devices is limited because of its high cost and lack of treatment planning tools. Hence, the MAASTRO applicator (for HDR 192Ir sources) was developed and characterized by measurements and Monte Carlo simulations to be a cost-effective alternative to CXB devices. METHODS AND MATERIALS: A cylindrical applicator with lateral shielding was designed to be used with a rectoscope using its tip as treatment surface. Both the applicator and the rectoscope have a slanted edge to potentially allow easier placement against tumors. The applicator design was achieved by Monte Carlo modeling and validated experimentally with film dosimetry, using the Papillon 50 (P50) device as reference. RESULTS: The applicator delivers CXB doses in less than 9 min using a 20375 U source for a treatment area of approximately 20 × 20 mm2 at 2 mm depth. Normalized at 2 mm, the dose falloff for depths of 0 mm, 5 mm, and 10 mm are 130%, 70%, and 43% for the P50 and 140%, 67%, and 38% for the MAASTRO applicator, respectively. CONCLUSIONS: The MAASTRO applicator was designed to use HDR 192Ir sources to deliver a dose distribution similar to those of CXB devices. The applicator may provide a cost-effective solution for endoluminal boosting with clinical treatment planning system integration.

A novel rectal applicator for contact radiotherapy with HDR 192Ir sources.

PURPOSE: Dose escalation to rectal tumors leads to higher complete response rates and may thereby enable omission of surgery. Important advantages of endoluminal boosting techniques include the possibility to apply a more selective/localized boost than using external beam radiotherapy. A novel brachytherapy (BT) rectal applicator with lateral shielding was designed to be used with a rectoscope for eye-guided positioning to deliver a dose distribution similar to the one of contact x-ray radiotherapy devices, using commonly available high-dose-rate 192Ir BT sources. METHODS AND MATERIALS: A cylindrical multichannel BT applicator with lateral shielding was designed by Monte Carlo modeling, validated experimentally with film dosimetry and compared with results found in the literature for the Papillon 50 (P50) contact x-ray radiotherapy device regarding rectoscope dimensions, radiation beam shape, dose fall-off, and treatment time. RESULTS: The multichannel applicator designed is able to deliver 30 Gy under 13 min with a 20350 U (5 Ci) source. The use of multiple channels and lateral shielding provide a uniform circular treatment surface with 22 mm in diameter. The resulting dose fall-off is slightly steeper (maximum difference of 5%) than the one generated by the P50 device with the 22 mm applicator. CONCLUSIONS: A novel multichannel rectal applicator for contact radiotherapy with high-dose-rate 192Ir sources that can be integrated with commercially available treatment planning systems was designed to produce a dose distribution similar to the one obtained by the P50 device.

Theoretical versus Ex Vivo Assessment of Radiation Damage Repair: An Investigation in Normal Breast Tissue.

In vivo validation of models of DNA damage repair will enable their use for optimizing clinical radiotherapy. In this study, a theoretical assessment was made of DNA double-strand break (DSB) induction in normal breast tissue after intraoperative radiation therapy (IORT), which is now an accepted form of adjuvant radiotherapy for selected patients with early breast cancer. DSB rates and relative biological effectiveness (RBE) were calculated as a function of dose, radiation quality and dose rate, each varying based on the applicator size used during IORT. The spectra of primary electrons in breast tissue adjacent to each applicator were calculated using measured X-ray spectra and Monte Carlo methods, and were used to inform a Monte Carlo damage simulation code. In the absence of repair, asymptotic RBE values (relative to (60)Co) were approximately 1.5. Beam-quality changes led to only minor variations in RBE among applicators, though differences in dose rate and overall dose delivery time led to larger variations and a rapid decrease in RBE. An experimental assessment of DSB induction was performed ex vivo using pre- and postirradiation tissue samples from patients receiving breast intraoperative radiation therapy. Relative DSB rates were assessed via γ-H2AX immunohistochemistry using proportional staining. Maximum-likelihood parameter estimation yielded a DSB repair halftime of 25.9 min (95% CI, 21.5-30.4 min), although the resulting model was not statistically distinguishable from one where there was no change in DSB yield among patients. Although the model yielded an in vivo repair halftime of the order of previous estimates for in vitro repair halftimes, we cannot conclude that it is valid in this context. This study highlights some of the uncertainties inherent in population analysis of ex vivo samples, and of the quantitative limitations of immunohistochemistry for assessment of DSB repair.

A comparison of the relative biological effectiveness of low energy electronic brachytherapy sources in breast tissue: a Monte Carlo study.

Electronic brachytherapy sources use low energy photons to treat the tumor bed during or after breast-conserving surgery. The relative biological effectiveness of two electronic brachytherapy sources was explored to determine if spectral differences due to source design influenced radiation quality and if radiation quality decreased with distance in the breast. The RBE was calculated through the number of DNA double strand breaks (RBEDSB) using the Monte Carlo damage simulator (MCDS) in combination with other Monte Carlo electron/photon spectrum calculations. 50kVp photons from the Intrabeam (Carl Zeiss Surgical) and Axxent (Xoft) through 40-mm spherical applicators were simulated to account for applicator and tissue attenuation in a variety of breast tissue compositions. 40kVp Axxent photons were also simulated. Secondary electrons (known to be responsible for most DNA damage) spectra at different distance were inputted into MCDS to calculate the RBEDSB. All RBEDSB used a cobalt-60 reference. RBEDSB data was combined with corresponding average photon spectrum energy for the Axxent and applied to model-based average photon energy distributions to produce an RBEDSB map of an accelerated partial breast irradiation (APBI) patient. Both Axxent and Intrabeam 50kVp spectra were shown to have a comparable RBEDSB of between 1.4 and 1.6 at all distances in spite of progressive beam hardening. The Axxent 40kVp also demonstrated a similar RBEDSB at distances. Most RBEDSB variability was dependent on the tissue type as was seen in rib (RBEDSB ≈ 1.4), gland (≈1.55), adipose (≈1.59), skin (≈1.52) and lung (≈1.50). RBEDSB variability between both sources was within 2%. A correlation was shown between RBEDSB and average photon energy and used to produce an RBEDSB map of a dose distribution in an APBI patient dataset. Radiation quality is very similar between electronic brachytherapy sources studied. No significant reductions in RBEDSB were observed with increasing distance from the source.

Preclinical Assessment of Efficacy of Radiation Dose Painting Based on Intratumoral FDG-PET Uptake.

PURPOSE: We tested therapeutic efficacy of two dose painting strategies of applying higher radiation dose to tumor subvolumes with high FDG uptake (biologic target volume, BTV): dose escalation and dose redistribution. We also investigated whether tumor response was determined by the highest dose in BTV or the lowest dose in gross tumor volume (GTV). EXPERIMENTAL DESIGN: FDG uptake was evaluated in rat rhabdomyosarcomas prior to irradiation. BTV was defined as 30% of GTV with the highest (BTVhot) or lowest (BTVcold) uptake. To test efficacy of dose escalation, tumor response (time to reach two times starting tumor volume, TGTV2) to Hot Boost irradiation (40% higher dose to BTVhot) was compared with Cold Boost (40% higher dose to BTVcold), while mean dose to GTV remained 12 Gy. To test efficacy of dose redistribution, TGTV2 after Hot Boost was compared with uniform irradiation with the same mean dose (8 or 12 Gy). RESULTS: TGTV2 after 12 Gy delivered heterogeneously (Hot and Cold Boost) or uniformly were not significantly different: 20.2, 19.5, and 20.6 days, respectively. Dose redistribution (Hot Boost) with 8 Gy resulted in faster tumor regrowth as compared with uniform irradiation (13.3 vs. 17.1 days; P = 0.026). Further increase in dose gradient to 60% led to a more pronounced decrease in TGTV2 (10.9 days; P < 0.0001). CONCLUSIONS: Dose escalation effect was independent of FDG uptake in target tumor volume, while dose redistribution was detrimental in this tumor model for dose levels applied here. Our data are consistent with the hypothesis that tumor response depends on the minimum intratumoral dose.

HDR 192Ir source speed measurements using a high speed video camera.

PURPOSE: The dose delivered with a HDR (192)Ir afterloader can be separated into a dwell component, and a transit component resulting from the source movement. The transit component is directly dependent on the source speed profile and it is the goal of this study to measure accurate source speed profiles. METHODS: A high speed video camera was used to record the movement of a (192)Ir source (Nucletron, an Elekta company, Stockholm, Sweden) for interdwell distances of 0.25-5 cm with dwell times of 0.1, 1, and 2 s. Transit dose distributions were calculated using a Monte Carlo code simulating the source movement. RESULTS: The source stops at each dwell position oscillating around the desired position for a duration up to (0.026 ± 0.005) s. The source speed profile shows variations between 0 and 81 cm/s with average speed of ∼ 33 cm/s for most of the interdwell distances. The source stops for up to (0.005 ± 0.001) s at nonprogrammed positions in between two programmed dwell positions. The dwell time correction applied by the manufacturer compensates the transit dose between the dwell positions leading to a maximum overdose of 41 mGy for the considered cases and assuming an air-kerma strength of 48 000 U. The transit dose component is not uniformly distributed leading to over and underdoses, which is within 1.4% for commonly prescribed doses (3-10 Gy). CONCLUSIONS: The source maintains its speed even for the short interdwell distances. Dose variations due to the transit dose component are much lower than the prescribed treatment doses for brachytherapy, although transit dose component should be evaluated individually for clinical cases.

The use of tetrahedral mesh geometries in Monte Carlo simulation of applicator based brachytherapy dose distributions.

Accounting for brachytherapy applicator attenuation is part of the recommendations from the recent report of AAPM Task Group 186. To do so, model based dose calculation algorithms require accurate modelling of the applicator geometry. This can be non-trivial in the case of irregularly shaped applicators such as the Fletcher Williamson gynaecological applicator or balloon applicators with possibly irregular shapes employed in accelerated partial breast irradiation (APBI) performed using electronic brachytherapy sources (EBS). While many of these applicators can be modelled using constructive solid geometry (CSG), the latter may be difficult and time-consuming. Alternatively, these complex geometries can be modelled using tessellated geometries such as tetrahedral meshes (mesh geometries (MG)). Recent versions of Monte Carlo (MC) codes Geant4 and MCNP6 allow for the use of MG. The goal of this work was to model a series of applicators relevant to brachytherapy using MG. Applicators designed for (192)Ir sources and 50 kV EBS were studied; a shielded vaginal applicator, a shielded Fletcher Williamson applicator and an APBI balloon applicator. All applicators were modelled in Geant4 and MCNP6 using MG and CSG for dose calculations. CSG derived dose distributions were considered as reference and used to validate MG models by comparing dose distribution ratios. In general agreement within 1% for the dose calculations was observed for all applicators between MG and CSG and between codes when considering volumes inside the 25% isodose surface. When compared to CSG, MG required longer computation times by a factor of at least 2 for MC simulations using the same code. MCNP6 calculation times were more than ten times shorter than Geant4 in some cases. In conclusion we presented methods allowing for high fidelity modelling with results equivalent to CSG. To the best of our knowledge MG offers the most accurate representation of an irregular APBI balloon applicator.

A medical image-based graphical platform -- features, applications and relevance for brachytherapy.

PURPOSE: Brachytherapy dose calculation is commonly performed using the Task Group-No 43 Report-Updated protocol (TG-43U1) formalism. Recently, a more accurate approach has been proposed that can handle tissue composition, tissue density, body shape, applicator geometry, and dose reporting either in media or water. Some model-based dose calculation algorithms are based on Monte Carlo (MC) simulations. This work presents a software platform capable of processing medical images and treatment plans, and preparing the required input data for MC simulations. METHODS AND MATERIALS: The A Medical Image-based Graphical platfOrm-Brachytherapy module (AMIGOBrachy) is a user interface, coupled to the MCNP6 MC code, for absorbed dose calculations. The AMIGOBrachy was first validated in water for a high-dose-rate (192)Ir source. Next, dose distributions were validated in uniform phantoms consisting of different materials. Finally, dose distributions were obtained in patient geometries. Results were compared against a treatment planning system including a linear Boltzmann transport equation (LBTE) solver capable of handling nonwater heterogeneities. RESULTS: The TG-43U1 source parameters are in good agreement with literature with more than 90% of anisotropy values within 1%. No significant dependence on the tissue composition was observed comparing MC results against an LBTE solver. Clinical cases showed differences up to 25%, when comparing MC results against TG-43U1. About 92% of the voxels exhibited dose differences lower than 2% when comparing MC results against an LBTE solver. CONCLUSION: The AMIGOBrachy can improve the accuracy of the TG-43U1 dose calculation by using a more accurate MC dose calculation algorithm. The AMIGOBrachy can be incorporated in clinical practice via a user-friendly graphical interface.

Comparison of TG-43 and TG-186 in breast irradiation using a low energy electronic brachytherapy source.

PURPOSE: The recently updated guidelines for dosimetry in brachytherapy in TG-186 have recommended the use of model-based dosimetry calculations as a replacement for TG-43. TG-186 highlights shortcomings in the water-based approach in TG-43, particularly for low energy brachytherapy sources. The Xoft Axxent is a low energy (<50 kV) brachytherapy system used in accelerated partial breast irradiation (APBI). Breast tissue is a heterogeneous tissue in terms of density and composition. Dosimetric calculations of seven APBI patients treated with Axxent were made using a model-based Monte Carlo platform for a number of tissue models and dose reporting methods and compared to TG-43 based plans. METHODS: A model of the Axxent source, the S700, was created and validated against experimental data. CT scans of the patients were used to create realistic multi-tissue/heterogeneous models with breast tissue segmented using a published technique. Alternative water models were used to isolate the influence of tissue heterogeneity and backscatter on the dose distribution. Dose calculations were performed using Geant4 according to the original treatment parameters. The effect of the Axxent balloon applicator used in APBI which could not be modeled in the CT-based model, was modeled using a novel technique that utilizes CAD-based geometries. These techniques were validated experimentally. Results were calculated using two dose reporting methods, dose to water (Dw,m) and dose to medium (Dm,m), for the heterogeneous simulations. All results were compared against TG-43-based dose distributions and evaluated using dose ratio maps and DVH metrics. Changes in skin and PTV dose were highlighted. RESULTS: All simulated heterogeneous models showed a reduced dose to the DVH metrics that is dependent on the method of dose reporting and patient geometry. Based on a prescription dose of 34 Gy, the average D90 to PTV was reduced by between ~4% and ~40%, depending on the scoring method, compared to the TG-43 result. Peak skin dose is also reduced by 10%-15% due to the absence of backscatter not accounted for in TG-43. The balloon applicator also contributed to the reduced dose. Other ROIs showed a difference depending on the method of dose reporting. CONCLUSIONS: TG-186-based calculations produce results that are different from TG-43 for the Axxent source. The differences depend strongly on the method of dose reporting. This study highlights the importance of backscatter to peak skin dose. Tissue heterogeneities, applicator, and patient geometries demonstrate the need for a more robust dose calculation method for low energy brachytherapy sources.

Dose perturbation due to catheter materials in high-dose-rate interstitial (192)Ir brachytherapy.

PURPOSE: Catheters made of either metal or plastic are currently used in brachytherapy treatment to insert radiative sources into patients. However, the radiation dose perturbations due to catheter attenuation are not taken into account in treatment planning. The purpose of this work is to quantify the effects of catheter composition on dose distribution and study their impacts on the overall treatment with high-dose-rate (192)Ir sources. METHODS AND MATERIALS: Dose perturbations are first studied in a simplified case consisting of two parallel catheters. The catheter wall is either composed of stainless steel or polyoxymethylene. The attenuations are studied as the distance between the two catheters is varied from 5 to 30 mm. Dose perturbations resulting from irradiation are evaluated with a Monte Carlo GEANT4 dose calculation algorithm. The dose differences are further investigated with seven typical high-dose-rate prostate treatment plans involving 17 catheters. RESULTS: The dose differences compared with water in the simplified case reach -4.3 ± 0.1% for stainless steel and 1.7 ± 0.5% for polyoxymethylene at 10 mm above the source when the catheters are separated by a distance of 5 mm. Dose perturbations are reduced in real treatment plans because of the contributions of the many dwell positions. Stainless steel and polyoxymethylene catheters induce on an average a dose difference of -1.3 ± 0.3% and 0.1 ± 0.2%, respectively in the target. CONCLUSIONS: The dose differences reported in this work do not warrant any changes in the clinical procedures.

Influence of trace elements in human tissue in low-energy photon brachytherapy dosimetry.

The aim of this paper is to determine the dosimetric impact of trace elements in human tissues for low-energy photon sources used in brachytherapy. Monte Carlo dose calculations were used to investigate the dosimetric effect of trace elements present in normal or cancerous human tissues. The effect of individual traces (atomic number Z = 11-30) was studied in soft tissue irradiated by low-energy brachytherapy sources. Three other tissue types (prostate, adipose and mammary gland) were also simulated with varying trace concentrations to quantify the contribution of each trace to the dose distribution. The dose differences between cancerous and healthy prostate tissues were calculated in single- and multi-source geometries. The presence of traces in a tissue produces a difference in the dose distribution that is dependent on Z and the concentration of the trace. Low-Z traces (Na) have a negligible effect (<0.3%) in all tissues, while higher Z (K) had a larger effect (>3%). There is a potentially significant difference in the dose distribution between cancerous and healthy prostate tissues (4%) and even larger if compared to the trace-free composition (15%) in both single- and multi-sourced geometries. Trace elements have a non-negligible (up to 8% in prostate D(90)) effect on the dose in tissues irradiated with low-energy photon sources. This study underlines the need for further investigation into accurate determination of the trace composition of tissues associated with low-energy brachytherapy. Alternatively, trace elements could be incorporated as a source of uncertainty in dose calculations.

Dose reduction in LDR brachytherapy by implanted prostate gold fiducial markers.

PURPOSE: The dosimetric impact of gold fiducial markers (FM) implanted prior to external beam radiotherapy of prostate cancer on low dose rate (LDR) brachytherapy seed implants performed in the context of combined therapy was investigated. METHODS: A virtual water phantom was designed containing a single FM. Single and multi source scenarios were investigated by performing Monte Carlo dose calculations, along with the influence of varying orientation and distance of the FM with respect to the sources. Three prostate cancer patients treated with LDR brachytherapy for a recurrence following external beam radiotherapy with implanted FM were studied as surrogate cases to combined therapy. FM and brachytherapy seeds were identified on post implant CT scans and Monte Carlo dose calculations were performed with and without FM. The dosimetric impact of the FM was evaluated by quantifying the amplitude of dose shadows and the volume of cold spots. D(90) was reported based on the post implant CT prostate contour. RESULTS: Large shadows are observed in the single source-FM scenarios. As expected from geometric considerations, the shadows are dependent on source-FM distance and orientation. Large dose reductions are observed at the distal side of FM, while at the proximal side a dose enhancement is observed. In multisource scenarios, the importance of shadows appears mitigated, although FM at the periphery of the seed distribution caused underdosage (

Consequences of dose heterogeneity on the biological efficiency of ¹°³Pd permanent breast seed implants.

Brachytherapy is associated with highly heterogeneous spatial dose distributions. This heterogeneity is usually ignored when estimating the biological effective dose (BED). In addition, the heterogeneities of the medium including the tissue heterogeneity (TH) and the interseed attenuation (ISA) are also contributing to the heterogeneity of the dose distribution, but they are both ignored in Task Group 43 (TG43)-based protocols. This study investigates the effect of dose heterogeneity, TH and ISA on metrics that are commonly used to quantify biological efficiency in brachytherapy. The special case of 29 breast cancer patients treated with permanent (103)Pd seed implant is considered here. BED is compared to equivalent uniform BED (EUBED) capable of considering the spatial heterogeneity of the dose distribution. The effects of TH and ISA on biological efficiency of treatments are taken into account by comparing TG43 with Monte Carlo (MC) dose calculations for each patient. The effect of clonogenic repopulation is also considered. The analysis is performed for different sets of (α/ß, α) ratios of (2, 0.3), (4, 0.27) and (10, 0.3) [Gy, Gy(-1)] covering the whole range of reported α/ß values in the literature. BED is sometimes larger and sometimes smaller than EUBED(TG43) indicating that the effect of the dose heterogeneity is not similar among patients. The effect of the dose heterogeneity can be characterized by using the D(99) dose metric. For each set of the radiobiological parameters considered, a D(99) threshold is found over which dose heterogeneity will cause an overestimation of the biological efficiencies while the inverse happens for smaller D(99) values. EUBED(MC) is always larger than EUBED(TG43) indicating that by neglecting TH and ISA in TG43-based dosimetry algorithms, the biological efficiencies may be underestimated by about 10 Gy. Overall, by going from BED to the more accurate EUBED(MC) there is a gain of about 9.6 to 13 Gy on the biological efficiency. The efficiency gain is about 10.8 to 14 Gy when the repopulation is considered. Dose heterogeneity does not have a constant impact on the biological efficiencies and may under- or overestimate the efficacy in different patients. However, the combined effect of neglecting dose heterogeneity, TH and ISA results in underestimation of the biological efficiencies in permanent breast seed implants.

Tissue modeling schemes in low energy breast brachytherapy.

Breast tissue is heterogeneous and is mainly composed of glandular (G) and adipose (A) tissues. The proportion of G versus A varies considerably among the population. The absorbed dose distributions in accelerated partial breast irradiation therapy with low energy photon brachytherapy sources are very sensitive to tissue heterogeneities. Current clinical algorithms use the recommendations of the AAPM TG43 report which approximates the human tissues by unit density water. The aim of this study is to investigate various breast tissue modeling schemes for low energy brachytherapy. A special case of breast permanent seed implant is considered here. Six modeling schemes are considered. Uniform and non-uniform water breast (UWB and NUWB) consider the density but neglect the effect of the composition of tissues. The uniform and the non-uniform G/A breast (UGAB and NUGAB) as well the age-dependent breast (ADB) models consider the effect of the composition. The segmented breast tissue (SBT) method uses a density threshold to distinguish between G and A tissues. The PTV D(90) metric is used for the analysis and is based on the dose to water (D(90(w,m))). D(90(m,m)) is also reported for comparison to D(90(w,m)). The two-month post-implant D(90(w,m)) averaged over 38 patients is smaller in NUWB than in UWB by about 4.6% on average (ranging from 5% to 13%). Large average differences of G/A breast models with TG43 (17% and 26% in UGAB and NUGAB, respectively) show that the effect of the chemical composition dominates the effect of the density on dose distributions. D(90(w,m)) is 12% larger in SBT than in TG43 when averaged. These differences can be as low as 4% or as high as 20% when the individual patients are considered. The high sensitivity of dosimetry on the modeling scheme argues in favor of an agreement on a standard tissue modeling approach to be used in low energy breast brachytherapy. SBT appears to generate the most geometrically reliable breast tissue models in this report.

Simulation study on potential accuracy gains from dual energy CT tissue segmentation for low-energy brachytherapy Monte Carlo dose calculations.

This work compares Monte Carlo (MC) dose calculations for (125)I and (103)Pd low-dose rate (LDR) brachytherapy sources performed in virtual phantoms containing a series of human soft tissues of interest for brachytherapy. The geometries are segmented (tissue type and density assignment) based on simulated single energy computed tomography (SECT) and dual energy (DECT) images, as well as the all-water TG-43 approach. Accuracy is evaluated by comparison to a reference MC dose calculation performed in the same phantoms, where each voxel's material properties are assigned with exactly known values. The objective is to assess potential dose calculation accuracy gains from DECT. A CT imaging simulation package, ImaSim, is used to generate CT images of calibration and dose calculation phantoms at 80, 120, and 140 kVp. From the high and low energy images electron density ρ(e) and atomic number Z are obtained using a DECT algorithm. Following a correction derived from scans of the calibration phantom, accuracy on Z and ρ(e) of ±1% is obtained for all soft tissues with atomic number Z ∊ [6,8] except lung. GEANT4 MC dose calculations based on DECT segmentation agreed with the reference within ±4% for (103)Pd, the most sensitive source to tissue misassignments. SECT segmentation with three tissue bins as well as the TG-43 approach showed inferior accuracy with errors of up to 20%. Using seven tissue bins in our SECT segmentation brought errors within ±10% for (103)Pd. In general (125)I dose calculations showed higher accuracy than (103)Pd. Simulated image noise was found to decrease DECT accuracy by 3-4%. Our findings suggest that DECT-based segmentation yields improved accuracy when compared to SECT segmentation with seven tissue bins in LDR brachytherapy dose calculation for the specific case of our non-anthropomorphic phantom. The validity of our conclusions for clinical geometry as well as the importance of image noise in the tissue segmentation procedure deserves further experimental investigation.

The difference of scoring dose to water or tissues in Monte Carlo dose calculations for low energy brachytherapy photon sources.

PURPOSE: The goal of this work is to compare D(m,m) (radiation transported in medium; dose scored in medium) and D(w,m) (radiation transported in medium; dose scored in water) obtained from Monte Carlo (MC) simulations for a subset of human tissues of interest in low energy photon brachytherapy. Using low dose rate seeds and an electronic brachytherapy source (EBS), the authors quantify the large cavity theory conversion factors required. The authors also assess whether ap plying large cavity theory utilizing the sources' initial photon spectra and average photon energy induces errors related to spatial spectral variations. First, ideal spherical geometries were investigated, followed by clinical brachytherapy LDR seed implants for breast and prostate cancer patients. METHODS: Two types of dose calculations are performed with the GEANT4 MC code. (1) For several human tissues, dose profiles are obtained in spherical geometries centered on four types of low energy brachytherapy sources: 125I, 103Pd, and 131Cs seeds, as well as an EBS operating at 50 kV. Ratios of D(w,m) over D(m,m) are evaluated in the 0-6 cm range. In addition to mean tissue composition, compositions corresponding to one standard deviation from the mean are also studied. (2) Four clinical breast (using 103Pd) and prostate (using 125I) brachytherapy seed implants are considered. MC dose calculations are performed based on postimplant CT scans using prostate and breast tissue compositions. PTV D90 values are compared for D(w,m) and D(m,m). RESULTS: (1) Differences (D(w,m)/D(m,m)-1) of -3% to 70% are observed for the investigated tissues. For a given tissue, D(w,m)/D(m,m) is similar for all sources within 4% and does not vary more than 2% with distance due to very moderate spectral shifts. Variations of tissue composition about the assumed mean composition influence the conversion factors up to 38%. (2) The ratio of D90(w,m) over D90(m,m) for clinical implants matches D(w,m)/D(m,m) at 1 cm from the single point sources, CONCLUSIONS: Given the small variation with distance, using conversion factors based on the emitted photon spectrum (or its mean energy) of a given source introduces minimal error. The large differences observed between scoring schemes underline the need for guidelines on choice of media for dose reporting. Providing such guidelines is beyond the scope of this work.

Sensitivity of low energy brachytherapy Monte Carlo dose calculations to uncertainties in human tissue composition.

PURPOSE: The objective of this work is to assess the sensitivity of Monte Carlo (MC) dose calculations to uncertainties in human tissue composition for a range of low photon energy brachytherapy sources: 125I, 103Pd, 131Cs, and an electronic brachytherapy source (EBS). The low energy photons emitted by these sources make the dosimetry sensitive to variations in tissue atomic number due to the dominance of the photoelectric effect. This work reports dose to a small mass of water in medium D(w,m) as opposed to dose to a small mass of medium in medium D(m,m). METHODS: Mean adipose, mammary gland, and breast tissues (as uniform mixture of the aforementioned tissues) are investigated as well as compositions corresponding to one standard deviation from the mean. Prostate mean compositions from three different literature sources are also investigated. Three sets of MC simulations are performed with the GEANT4 code: (1) Dose calculations for idealized TG-43-like spherical geometries using point sources. Radial dose profiles obtained in different media are compared to assess the influence of compositional uncertainties. (2) Dose calculations for four clinical prostate LDR brachytherapy permanent seed implants using 125I seeds (Model 2301, Best Medical, Springfield, VA). The effect of varying the prostate composition in the planning target volume (PTV) is investigated by comparing PTV D90 values. (3) Dose calculations for four clinical breast LDR brachytherapy permanent seed implants using 103Pd seeds (Model 2335, Best Medical). The effects of varying the adipose/gland ratio in the PTV and of varying the elemental composition of adipose and gland within one standard deviation of the assumed mean composition are investigated by comparing PTV D90 values. For (2) and (3), the influence of using the mass density from CT scans instead of unit mass density is also assessed. RESULTS: Results from simulation (1) show that variations in the mean compositions of tissues affect low energy brachytherapy dosimetry. Dose differences between mean and one standard deviation of the mean composition increasing with distance from the source are observed. It is established that the 125I and 131Cs sources are the least sensitive to variations in elemental compositions while 103Pd is most sensitive. The EBS falls in between and exhibits complex behavior due to significant spectral hardening. Results from simulation (2) show that two prostate compositions are dosimetrically equivalent to water while the third shows D90 differences of up to 4%. Results from simulation (3) show that breast is more sensitive than prostate with dose variations of up to 30% from water for 70% adipose/30% gland breast. The variability of the breast composition adds a +/- 10% dose variation. CONCLUSIONS: Low energy brachytherapy dose distributions in tissue differ from water and are influenced by density, mean tissue composition, and patient-to-patient composition variations. The results support the use of a dose calculation algorithm accounting for heterogeneities such as MC. Since this work shows that variations in mean tissue compositions affect MC dosimetry and result in increased dose uncertainties, the authors conclude that imaging tools providing more accurate estimates of elemental compositions such as dual energy CT would be beneficial.