J. G. H. Sutherland

Monte Carlo calculated absorbed-dose energy dependence of EBT and EBT2 film

PURPOSE The absorbed-dose energy dependence of GAFCHROMIC EBT and EBT2 film irradiated in photon beams is studied to understand the shape of the curves and the physics behind them. METHODS The absorbed-dose energy dependence is calculated using the EGSnrc-based EGS_chamber and DOSRZnrc codes by calculating the ratio of dose to water to dose to active film layers at photon energies ranging from 3 keV to 18 MeV. These data are compared to the mass energy absorption coefficient ratios and the restricted stopping power ratios of water to active film materials as well as to previous experimental results. RESULTS In the photon energy range of 100 keV to 18 MeV the absorbed-dose energy dependence is found to be energy independent within +/- 0.6%. However, below 100 keV, the absorbed-dose energy dependence of EBT varies by approximately 10% due to changes in mass energy absorption coefficient ratios of water to film materials, as well as an increase in the number of electrons being created and scattered in the central surface layer of the film. Results are found to disagree with previous experimental studies suggesting the possibility of an intrinsic energy dependence at lower photon energies. For EBT2 film the absorbed-dose energy dependence at low photon energies varies by 50% or 10% depending on the manufacturing lot due to changes in the ratio of mass energy absorption coefficients of the active emulsion layers to water. CONCLUSIONS Caution is recommended when using GAFCHROMIC EBT/EBT2 films at photon energies below 100 keV. It is recommended that the effective atomic number of future films be produced as close to that of water and that thicker active layers are advantageous.

Sci—Sat AM(2): Brachy — 05: Fast Monte Carlo Dose Calculations for Brachytherapy with BrachyDose

A fast dose calculation algorithm called BrachyDose has been developed for brachytherapy applications. BrachyDose is based on the EGSnrc code system for simulating radiation transport. Complex geometries are modelled through the superposition of basic geometric entities (spheres, cuboids, cylinders, and cones) using Yegins multi-geometry package; the phantom geometry may be defined using a CT dataset. A database of brachytherapy sources has been developed and benchmarked, as has a database of eye plaque applicators. BrachyDose scores collision kerma, which is equivalent to absorbed dose for most situations of interest, using a tracklength estimator. The phase space of particles emitted from brachytherapy sources may be generated with BrachyDose and used in subsequent simulations to avoid the repeated simulation of particle transport within sources. A particle recycling feature has been implemented for multisource configurations in which the first source acts as a particle generator; particles emitted from this source are reinitiated at each source location. Dose calculations for prostate permanent implants achieving 2% average uncertainty in the prostate region take less than 30 seconds in (2 mm)3 voxels on a single 3.0 GHz Woodcrest core; calculation times for eye plaque therapy are on the order of three minutes in (0.5 mm)3 voxels. These calculation times are sufficiently fast for routine clinical treatment planning. A graphical user interface (GUI) for BrachyDose has been developed. Working towards clinical implementation, efforts are underway to integrate data in the DICOM-RT format with BrachyDose.

Model-based dose calculations for 125I lung brachytherapy

PURPOSE Model-baseddose calculations (MBDCs) are performed using patient computed tomography (CT) data for patients treated with intraoperative (125)I lung brachytherapy at the Mayo Clinic Rochester. Various metallic artifact correction and tissue assignment schemes are considered and their effects on dose distributions are studied. Dose distributions are compared to those calculated under TG-43 assumptions. METHODS Dose distributions for six patients are calculated using phantoms derived from patient CT data and the EGSnrc user-code BrachyDose. (125)I (GE Healthcare/Oncura model 6711) seeds are fully modeled. Four metallic artifact correction schemes are applied to the CT data phantoms: (1) no correction, (2) a filtered back-projection on a modified virtual sinogram, (3) the reassignment of CT numbers above a threshold in the vicinity of the seeds, and (4) a combination of (2) and (3). Tissue assignment is based on voxel CT number and mass density is assigned using a CT number to mass density calibration. Three tissue assignment schemes with varying levels of detail (20, 11, and 5 tissues) are applied to metallic artifact corrected phantoms. Simulations are also performed under TG-43 assumptions, i.e., seeds in homogeneous water with no interseed attenuation. RESULTS Significant dose differences (up to 40% for D(90)) are observed between uncorrected and metallic artifact corrected phantoms. For phantoms created with metallic artifact correction schemes (3) and (4), dose volume metrics are generally in good agreement (less than 2% differences for all patients) although there are significant local dose differences. The application of the three tissue assignment schemes results in differences of up to 8% for D(90); these differences vary between patients. Significant dose differences are seen between fully modeled and TG-43 calculations with TG-43 underestimating the dose (up to 36% in D(90)) for larger volumes containing higher proportions of healthy lung tissue. CONCLUSIONS Metallic artifact correction is necessary for accurate application of MBDCs for lung brachytherapy; simpler threshold replacement methods may be sufficient for early adopters concerned with clinical dose metrics. Rigorous determination of voxel tissue parameters and tissue assignment is required for accurate dose calculations as different tissue assignment schemes can result in significantly different dose distributions. Significant differences are seen between MBDCs and TG-43 dose distributions with TG-43 underestimating dose in volumes containing healthy lung tissue.PURPOSE Model-baseddose calculations (MBDCs) are performed using patient computed tomography (CT) data for patients treated with intraoperative125 I lung brachytherapy at the Mayo Clinic Rochester. Various metallic artifact correction and tissue assignment schemes are considered and their effects on dose distributions are studied. Dose distributions are compared to those calculated under TG-43 assumptions. METHODS Dose distributions for six patients are calculated using phantoms derived from patient CT data and the EGSnrc user-code BrachyDose.125 I (GE Healthcare/Oncura model 6711) seeds are fully modeled. Four metallic artifact correction schemes are applied to the CT data phantoms: (1) no correction, (2) a filtered back-projection on a modified virtual sinogram, (3) the reassignment of CT numbers above a threshold in the vicinity of the seeds, and (4) a combination of (2) and (3). Tissue assignment is based on voxel CT number and mass density is assigned using a CT number to mass density calibration. Three tissue assignment schemes with varying levels of detail (20, 11, and 5 tissues) are applied to metallic artifact corrected phantoms. Simulations are also performed under TG-43 assumptions, i.e., seeds in homogeneous water with no interseed attenuation. RESULTS Significant dose differences (up to 40% for D90 ) are observed between uncorrected and metallic artifact corrected phantoms. For phantoms created with metallic artifact correction schemes (3) and (4), dose volume metrics are generally in good agreement (less than 2% differences for all patients) although there are significant local dose differences. The application of the three tissue assignment schemes results in differences of up to 8% for D90 ; these differences vary between patients. Significant dose differences are seen between fully modeled and TG-43 calculations with TG-43 underestimating the dose (up to 36% in D90 ) for larger volumes containing higher proportions of healthy lung tissue. CONCLUSIONS Metallic artifact correction is necessary for accurate application of MBDCs for lung brachytherapy; simpler threshold replacement methods may be sufficient for early adopters concerned with clinical dose metrics. Rigorous determination of voxel tissue parameters and tissue assignment is required for accurate dose calculations as different tissue assignment schemes can result in significantly different dose distributions. Significant differences are seen between MBDCs and TG-43 dose distributions with TG-43 underestimating dose in volumes containing healthy lung tissue.

Model-based dose calculations for {sup 125}I lung brachytherapy

PURPOSE Model-baseddose calculations (MBDCs) are performed using patient computed tomography (CT) data for patients treated with intraoperative (125)I lung brachytherapy at the Mayo Clinic Rochester. Various metallic artifact correction and tissue assignment schemes are considered and their effects on dose distributions are studied. Dose distributions are compared to those calculated under TG-43 assumptions. METHODS Dose distributions for six patients are calculated using phantoms derived from patient CT data and the EGSnrc user-code BrachyDose. (125)I (GE Healthcare/Oncura model 6711) seeds are fully modeled. Four metallic artifact correction schemes are applied to the CT data phantoms: (1) no correction, (2) a filtered back-projection on a modified virtual sinogram, (3) the reassignment of CT numbers above a threshold in the vicinity of the seeds, and (4) a combination of (2) and (3). Tissue assignment is based on voxel CT number and mass density is assigned using a CT number to mass density calibration. Three tissue assignment schemes with varying levels of detail (20, 11, and 5 tissues) are applied to metallic artifact corrected phantoms. Simulations are also performed under TG-43 assumptions, i.e., seeds in homogeneous water with no interseed attenuation. RESULTS Significant dose differences (up to 40% for D(90)) are observed between uncorrected and metallic artifact corrected phantoms. For phantoms created with metallic artifact correction schemes (3) and (4), dose volume metrics are generally in good agreement (less than 2% differences for all patients) although there are significant local dose differences. The application of the three tissue assignment schemes results in differences of up to 8% for D(90); these differences vary between patients. Significant dose differences are seen between fully modeled and TG-43 calculations with TG-43 underestimating the dose (up to 36% in D(90)) for larger volumes containing higher proportions of healthy lung tissue. CONCLUSIONS Metallic artifact correction is necessary for accurate application of MBDCs for lung brachytherapy; simpler threshold replacement methods may be sufficient for early adopters concerned with clinical dose metrics. Rigorous determination of voxel tissue parameters and tissue assignment is required for accurate dose calculations as different tissue assignment schemes can result in significantly different dose distributions. Significant differences are seen between MBDCs and TG-43 dose distributions with TG-43 underestimating dose in volumes containing healthy lung tissue.PURPOSE Model-baseddose calculations (MBDCs) are performed using patient computed tomography (CT) data for patients treated with intraoperative125 I lung brachytherapy at the Mayo Clinic Rochester. Various metallic artifact correction and tissue assignment schemes are considered and their effects on dose distributions are studied. Dose distributions are compared to those calculated under TG-43 assumptions. METHODS Dose distributions for six patients are calculated using phantoms derived from patient CT data and the EGSnrc user-code BrachyDose.125 I (GE Healthcare/Oncura model 6711) seeds are fully modeled. Four metallic artifact correction schemes are applied to the CT data phantoms: (1) no correction, (2) a filtered back-projection on a modified virtual sinogram, (3) the reassignment of CT numbers above a threshold in the vicinity of the seeds, and (4) a combination of (2) and (3). Tissue assignment is based on voxel CT number and mass density is assigned using a CT number to mass density calibration. Three tissue assignment schemes with varying levels of detail (20, 11, and 5 tissues) are applied to metallic artifact corrected phantoms. Simulations are also performed under TG-43 assumptions, i.e., seeds in homogeneous water with no interseed attenuation. RESULTS Significant dose differences (up to 40% for D90 ) are observed between uncorrected and metallic artifact corrected phantoms. For phantoms created with metallic artifact correction schemes (3) and (4), dose volume metrics are generally in good agreement (less than 2% differences for all patients) although there are significant local dose differences. The application of the three tissue assignment schemes results in differences of up to 8% for D90 ; these differences vary between patients. Significant dose differences are seen between fully modeled and TG-43 calculations with TG-43 underestimating the dose (up to 36% in D90 ) for larger volumes containing higher proportions of healthy lung tissue. CONCLUSIONS Metallic artifact correction is necessary for accurate application of MBDCs for lung brachytherapy; simpler threshold replacement methods may be sufficient for early adopters concerned with clinical dose metrics. Rigorous determination of voxel tissue parameters and tissue assignment is required for accurate dose calculations as different tissue assignment schemes can result in significantly different dose distributions. Significant differences are seen between MBDCs and TG-43 dose distributions with TG-43 underestimating dose in volumes containing healthy lung tissue.

Changes in dose with segmentation of breast tissues in Monte Carlo calculations for low-energy brachytherapy

PURPOSE To investigate the use of various breast tissue segmentation models in Monte Carlo dose calculations for low-energy brachytherapy. METHODS The EGSnrc user-code BrachyDose is used to perform Monte Carlo simulations of a breast brachytherapy treatment using TheraSeed Pd-103 seeds with various breast tissue segmentation models. Models used include a phantom where voxels are randomly assigned to be gland or adipose (randomly segmented), a phantom where a single tissue of averaged gland and adipose is present (averaged tissue), and a realistically segmented phantom created from previously published numerical phantoms. Radiation transport in averaged tissue while scoring in gland along with other combinations is investigated. The inclusion of calcifications in the breast is also studied in averaged tissue and randomly segmented phantoms. RESULTS In randomly segmented and averaged tissue phantoms, the photon energy fluence is approximately the same; however, differences occur in the dose volume histograms (DVHs) as a result of scoring in the different tissues (gland and adipose versus averaged tissue), whose mass energy absorption coefficients differ by 30%. A realistically segmented phantom is shown to significantly change the photon energy fluence compared to that in averaged tissue or randomly segmented phantoms. Despite this, resulting DVHs for the entire treatment volume agree reasonably because fluence differences are compensated by dose scoring differences. DVHs for the dose to only the gland voxels in a realistically segmented phantom do not agree with those for dose to gland in an averaged tissue phantom. Calcifications affect photon energy fluence to such a degree that the differences in fluence are not compensated for (as they are in the no calcification case) by dose scoring in averaged tissue phantoms. CONCLUSIONS For low-energy brachytherapy, if photon transport and dose scoring both occur in an averaged tissue, the resulting DVH for the entire treatment volume is reasonably accurate because inaccuracies in photon energy fluence are compensated for by inaccuracies in localized dose scoring. If dose to fibroglandular tissue in the breast is of interest, then the inaccurate photon energy fluence calculated in an averaged tissue phantom will result in inaccurate DVHs and average doses for those tissues. Including calcifications necessitates the use of proper tissue segmentation.

Metallic artifact mitigation and organ-constrained tissue assignment for Monte Carlo calculations of permanent implant lung brachytherapy.

PURPOSE To investigate methods of generating accurate patient-specific computational phantoms for the Monte Carlo calculation of lung brachytherapy patient dose distributions. METHODS Four metallic artifact mitigation methods are applied to six lung brachytherapy patient computed tomography (CT) images: simple threshold replacement (STR) identifies high CT values in the vicinity of the seeds and replaces them with estimated true values; fan beam virtual sinogram replaces artifact-affected values in a virtual sinogram and performs a filtered back-projection to generate a corrected image; 3D median filter replaces voxel values that differ from the median value in a region of interest surrounding the voxel and then applies a second filter to reduce noise; and a combination of fan beam virtual sinogram and STR. Computational phantoms are generated from artifact-corrected and uncorrected images using several tissue assignment schemes: both lung-contour constrained and unconstrained global schemes are considered. Voxel mass densities are assigned based on voxel CT number or using the nominal tissue mass densities. Dose distributions are calculated using the EGSnrc user-code BrachyDose for (125)I, (103)Pd, and (131)Cs seeds and are compared directly as well as through dose volume histograms and dose metrics for target volumes surrounding surgical sutures. RESULTS Metallic artifact mitigation techniques vary in ability to reduce artifacts while preserving tissue detail. Notably, images corrected with the fan beam virtual sinogram have reduced artifacts but residual artifacts near sources remain requiring additional use of STR; the 3D median filter removes artifacts but simultaneously removes detail in lung and bone. Doses vary considerably between computational phantoms with the largest differences arising from artifact-affected voxels assigned to bone in the vicinity of the seeds. Consequently, when metallic artifact reduction and constrained tissue assignment within lung contours are employed in generated phantoms, this erroneous assignment is reduced, generally resulting in higher doses. Lung-constrained tissue assignment also results in increased doses in regions of interest due to a reduction in the erroneous assignment of adipose to voxels within lung contours. Differences in dose metrics calculated for different computational phantoms are sensitive to radionuclide photon spectra with the largest differences for (103)Pd seeds and smallest but still considerable differences for (131)Cs seeds. CONCLUSIONS Despite producing differences in CT images, dose metrics calculated using the STR, fan beam + STR, and 3D median filter techniques produce similar dose metrics. Results suggest that the accuracy of dose distributions for permanent implant lung brachytherapy is improved by applying lung-constrained tissue assignment schemes to metallic artifact corrected images.

A Monte Carlo investigation of lung brachytherapy treatment planning

Iodine-125 ((125)I) and Caesium-131 ((131)Cs) brachytherapy have been used in conjunction with sublobar resection to reduce the local recurrence of stage I non-small cell lung cancer compared with resection alone. Treatment planning for this procedure is typically performed using only a seed activity nomogram or look-up table to determine seed strand spacing for the implanted mesh. Since the post-implant seed geometry is difficult to predict, the nomogram is calculated using the TG-43 formalism for seeds in a planar geometry. In this work, the EGSnrc user-code BrachyDose is used to recalculate nomograms using a variety of tissue models for (125)I and (131)Cs seeds. Calculated prescription doses are compared to those calculated using TG-43. Additionally, patient CT and contour data are used to generate virtual implants to study the effects that post-implant deformation and patient-specific tissue heterogeneity have on perturbing nomogram-derived dose distributions. Differences of up to 25% in calculated prescription dose are found between TG-43 and Monte Carlo calculations with the TG-43 formalism underestimating prescription doses in general. Differences between the TG-43 formalism and Monte Carlo calculated prescription doses are greater for (125)I than for (131)Cs seeds. Dose distributions are found to change significantly based on implant deformation and tissues surrounding implants for patient-specific virtual implants. Results suggest that accounting for seed grid deformation and the effects of non-water media, at least approximately, are likely required to reliably predict dose distributions in lung brachytherapy patients.

SU‐E‐T‐732: The Effects of Calcification Modeling for Monte Carlo Dose Calculations of Low Dose Rate Prostate Brachytherapy

Purpose: Retrospective Monte Carlo calculations recently revealed that tissue calcifications cause significant dose perturbations in low dose rate prostate brachytherapy patients. This work investigates dose differences due to calcification modeling schemes of varying complexity for MC dose calculations. Methods: A prostate cancer patient with a large prostatic calcification (∼2.3 cm3) treated with 125I seeds was studied. MC calculations were performed using the EGSnrc user-code BrachyDose with dose scored to medium in medium. Computational phantoms were generated from CT images acquired one month post-implant. Tissues were assigned to voxels within structure contours based on CT number: prostate and calcification within the target, air and muscle within the rectum, urinary bladder within the bladder, and ICRU 46 soft tissue and bone in remaining voxels. Target models included: water only (TG-43), prostate tissue with CT-derived mass densities except for higher calcification densities over¬ridden with prostate nominal density, only prostate with unmodified CT-derived densities, prostate and 100% calcification with CT-derived densities (PC), and prostate and incremental mixtures of prostate and calcification with CT-derived densities (PCmix). Results: Local dose differences of up to a factor of 3 or more were found between the various models including significant dose differences between PC and PCmix inside and adjacent to the calcification. Dose metrics differ between the models with the minimum dose that 90% of the target received (D90) differing by 4% between PC and PCmix. Adjacent to calcification, doses for calcification models were approximately 50% less than those calculated by TG-43. D90 for calcification models was approximately 80–90% the value of that calculated by TG-43. Conclusions: Considerable dose differences were found between calculations using various calcification modeling schemes for a low dose rate prostate brachytherapy patient with prostatic calcifications, highlighting the importance of detailed and accurate calcification modeling for MC dose calculations of these treatments.

WE‐C‐108‐11: Patient‐Specific Monte Carlo Simulation of Lung Brachytherapy: Metallic Artifact Reduction and Organ‐Constrained Tissue Assignment

PURPOSE To investigate techniques for developing accurate patient-specific computational phantoms for Monte Carlo simulations of lung brachytherapy. Methods to mitigate streaking artifacts due to brachytherapy sources in CT images and organ-constrained tissue assignment schemes are explored. METHODS Three different metallic artifact reduction (MAR) techniques are applied to lung brachytherapy patient CT data: thresholding replaces high CT values with estimated true values; fan-beam virtual sinogram replaces artifact-affected values in a virtual sinogram; and 3D median filter uses local voxel values to guide the replacement of outlying voxel values. Computational phantoms are generated from metallic artifact corrected and uncorrected images with voxel composition defined by CT number and density defined either by nominal tissue density or a CT number to density calibration. Multiple tissue assignment schemes are considered, including some with organ-specific constraints. Dose distributions for I-125 and Cs-131 seeds are calculated using the EGSnrc user-code BrachyDose and are compared directly as well as through DVHs and dose metrics for target volumes surrounding surgical sutures. RESULTS The most effective MAR technique is the virtual sinogram method as it effectively mitigates artifacts while maintaining image integrity. Thresholding only removes artifacts in the vicinity of sources (leaving artifacts in other regions) and median filtering degrades tissue heterogeneities. Dose distributions for phantoms treated with various MAR techniques and tissue assignment schemes can vary significantly. Dose differences between MAR and uncorrected phantoms are reduced by constraining tissue assignments in lung contours; in one example, differences in D90 of 20% are reduced to 3%. Phantoms with nominal densities Result in higher doses than those with CT derived densities. CONCLUSION Application of MAR techniques to CT data is necessary to develop realistic computational phantoms for Monte Carlo simulations; the virtual sinogram method is a promising image-based technique. Organ-constrained tissue assignment is necessary for accurate tissue assignment. This work was supported by the Natural Sciences and Engineering Research Council, the Canada Research Chairs program, and an Ontario Graduate Scholarship.

Sci—Sat AM: Brachy — 11: Improving treatment planning for I‐125 lung brachytherapy using Monte Carlo methods

125 I brachytherapy used in conjunction with sublobar resection to treat stage I non-small cell lung cancer has been reported to improve disease-free and overall survival rates compared with resection alone. Treatments are planned intra-operatively using seed spacing nomograms or tables to achieve a prescription dose defined 5 mm above the implant plane. Dose distributions for patients treated with this technique at the Mayo Clinic Rochester were reanalyzed using a Monte Carlo (MC) calculation; significant differences were observed between the standard TG-43 dose calculations and the actual dose delivered as determined by MC. This work investigates differences between TG-43 calculated prescription doses and those calculated in more accurate models. Monte Carlo calculations are performed using the EGSnrc user-code BrachyDose with a number of lung tissue phantom models including patient CT-derived phantoms. Seed spacing nomograms using these models are recalculated by determining the dose to the prescription point using the activities per seed required to produce a prescription dose of 100 Gy with the TG-43 point source formalism. Models using nominal density lung or CT-derived density lung tissue result in a significant increase in dose to the prescription point (up to approximately 25%) compared to TG-43 calculated doses. The differences observed suggest that patients routinely receive significantly higher doses than planned using TG-43 derived nomograms. Additionally, deviation from TG-43 increases as seed spacing increases. Media heterogeneities significantly affect dose distributions and prescription doses for 125 I lung brachytherapy, underlining the importance of using model-based dose calculation algorithms to plan and analyze these treatments.