Biomedical subjects

D W O Rogers

Publications and source records attributed to D W O Rogers.

At least 19 recordsLinked to original sources

PubMed2006

Fifty years of Monte Carlo simulations for medical physics.

Monte Carlo techniques have become ubiquitous in medical physics over the last 50 years with a doubling of papers on the subject every 5 years between the first PMB paper in 1967 and 2000 when the numbers levelled off. While recognizing the many other roles that Monte Carlo techniques have played in medical physics, this review emphasizes techniques for electron-photon transport simulations. The broad range of codes available is mentioned but there is special emphasis on the EGS4/EGSnrc code system which the author has helped develop for 25 years. The importance of the 1987 Erice Summer School on Monte Carlo techniques is highlighted. As an illustrative example of the role Monte Carlo techniques have played, the history of the correction for wall attenuation and scatter in an ion chamber is presented as it demonstrates the interplay between a specific problem and the development of tools to solve the problem which in turn leads to applications in other areas.

Animals↗

PubMed2006

Wall correction factors, Pwall, for thimble ionization chambers.

The EGSnrc Monte Carlo user-code CSnrc is used to calculate wall correction factors, Pwall, for thimble ionization chambers in photon and electron beams. CSnrc calculated values of Pwall give closer agreement with previous experimental results than do the values from the standard formalism used in current dosimetry protocols. A set of Pwall values, computed at the reference depth in water, is presented for several commonly used thimble chambers. These values differ from the commonly used values by up to 0.8% for megavoltage photon beams, particularly for nominal beam energies below 6 MV. The sleeve effect, which is not currently taken into account by the TG-51 dosimetry protocol, is computed to be up to 0.3% and is in some cases larger than the Pwal1 correction itself. In electron beams, where dosimetry protocols assume a wall correction of unity, CSnrc calculations show Pwall values of up to 0.6% at the reference depth, depending on the wall material. Pwall is shown to be sensitive to the depth of measurement, varying by 2.5% for a graphite-walled cylindrical Farmer-like chamber between a depth of 0.5 cm and R50 in a 6 MeV electron beam.

Algorithms↗

PubMed2006

Wall correction factors, Pwall, for parallel-plate ionization chambers.

The EGSnrc Monte Carlo user-code CSnrc is used to calculate wall correction factors, Pwall,, for parallel-plate ionization chambers in photon and electron beams. A set of Pwall values, computed at the reference depth in water, is presented for several commonly used parallel-plate chambers. These values differ from the standard assumption of unity used by dosimetry protocols by up to 1.7% for clinical electron beams. Calculations also show that Pwall is strongly dependent on the depth of measurement and can vary by as much as 6% for a 6 MeV beam in moving from a depth of dref to a depth of R50. In photon beams, where there is limited information available regarding Pwall for parallel-plate chambers, CSnrc calculations show Pwall values of up to 2.4% at the reference depth over a range of photon energies. The Pwall values for photon beams are in good agreement with previous estimates of the wall correction but have much lower statistical uncertainties and cover a wider range of photon beam energies.

Algorithms↗

PubMed2006

On the accuracy of techniques for obtaining the calibration coefficient N(K) of 192Ir HDR brachytherapy sources.

The accuracy of interpolation or averaging procedures for obtaining the calibration coefficient N(K) for 192Ir high-dose-rate brachytherapy sources has been investigated using the EGSnrc Monte Carlo simulation system. It is shown that the widely used two-point averaging procedure of Goetsch et al. [Med. Phys. 18, 462 (1991)] has some conceptual problems. Most importantly, they recommended, as did the IAEA, averaging A(wall)N(K) values whereas one should average 1/N(K) values. In practice this and other issues are shown to have little effect except for Goetsch et al.'s methods for determining A(wall) values. Their method of generalizing the A(wall) values measured in one geometry to other geometries is incorrect by up to 2%. However, these errors in A(wall) values cause systematic errors of only 0.3% in 192Ir calibration coefficients. It is shown that A(wall) values need not be included in the averaging technique at all, thereby simplifying the technique considerably. It is demonstrated that as long as ion chambers with a flat response are used and/or very heavily filtered 250 kV (or higher) beams of x rays are used in the averaging, then almost all techniques can provide adequate accuracy.

Brachytherapy↗

PubMed2005

An experimental and Monte Carlo investigation of the energy dependence of alanine/EPR dosimetry: II. Clinical electron beams.

The energy dependence of alanine/EPR dosimetry for 8, 12, 18 and 22 MeV clinical electron beams was investigated by experiment and by Monte Carlo simulations. Alanine pellets in a waterproof holder were irradiated in a water phantom using an Elekta Precise linear accelerator. The dose rates at the reference point were determined following the TG-51 protocol using an NACP-02 parallel-plate chamber calibrated in a (60)Co beam. The EPR spectra of irradiated pellets were measured using a Bruker EMX 081 EPR spectrometer. Experimentally, we found no significant change in alanine/EPR response to absorbed dose-to-water over the energy range 8-22 MeV at an uncertainty level of 0.6%. However, the response for high-energy electrons is about 1.3 (+/-1.1)% lower than for (60)Co. The EGSnrc Monte Carlo system was used to calculate the ratio of absorbed dose-to-alanine to absorbed dose-to-water and it was shown that there is 1.3 (+/-0.2)% reduction in this ratio from the (60)Co beam to the electron beams, which confirms the experimental results. Alanine/EPR response per unit absorbed dose-to-alanine was also investigated and it is the same for high-energy electrons and (60)Co gamma-rays.

Alanine↗

PubMed2005

Calculation of photon energy deposition kernels and electron dose point kernels in water.

Effects of changes in the physics of EGSnrc compared to EGS4/PRESTA on energy deposition kernels for monoenergetic photons and on dose point kernels for beta sources in water are investigated. In the diagnostic energy range, Compton binding corrections were found to increase the primary energy fraction up to 4.5% at 30 keV with a corresponding reduction of the scatter component of the kernels. Rayleigh scattered photons significantly increase the scatter component of the kernels and reduce the primary energy fraction with a maximum 12% reduction also at 30 keV where the Rayleigh cross section in water has its maximum value. Sampling the photo-electron angular distribution produces a redistribution of the energy deposited by primaries around the interaction site causing differences of up to 2.7 times in the backscattered energy fraction at 20 keV. Above the pair production threshold, the dose distribution versus angle of the primary dose component is significantly different from the EGS4 results. This is related to the more accurate angular sampling of the electron-positron pair direction in EGSnrc as opposed to using a fixed angle approximation in default EGS4. Total energy fractions for photon beams obtained with EGSnrc and EGS4 are almost the same within 0.2%. This fact suggests that the estimate of the total dose at a given point inside an infinite homogeneous water phantom irradiated by broad beams of photons will be very similar for kernels calculated with both codes. However, at interfaces or near boundaries results can be very different especially in the diagnostic energy range. EGSnrc calculated kernels for monoenergetic electrons (50 keV, 100 keV, and 1 MeV) and beta spectra (32P and 90Y) are in excellent agreement with reported EGS4 values except at 1 MeV where inclusion of spin effects in EGSnrc produces an increase of the effective range of electrons. Comparison at 1 MeV with an ETRAN calculation of the electron dose point kernel shows excellent agreement.

Algorithms↗

PubMed2005

A Monte Carlo derived TG-51 equivalent calibration for helical tomotherapy.

Helical tomotherapy (HT) requires a method of accurately determining the absorbed dose under reference conditions. In the AAPM's TG-51 external beam dosimetry protocol, the quality conversion factor, kQ, is presented as a function of the photon component of the percentage depth-dose at 10 cm depth, %dd(10)x, measured under the reference conditions of a 10 x 10 cm2 field size and a source-to-surface distance (SSD) of 100 cm. The value of %dd(10)x from HT cannot be used for the determination of kQ because the design of the HT does not meet the following TG-51 reference conditions: (i) the field size and the practical SSD required by TG-51 are not obtainable and (ii) the absence of the flattening filter changes the beam quality thus affecting some components of kQ. The stopping power ratio is not affected because of its direct relationship to %dd(10)x. We derive a relationship for the Exradin A1SL ion chamber converting the %dd(10)x measured under HT "reference conditions" of SSD=85 cm and a 5 x 10 cm2 field-size [%dd(10)x[HT Ref]], to the dosimetric equivalent value under for TG-51 reference conditions [%dd(10)x[HT TG-51]] for HT. This allows the determination of kQ under the HT reference conditions. The conversion results in changes of 0.1% in the value of kQ for our particular unit. The conversion relationship should also apply to other ion chambers with possible errors on the order of 0.1%.

Algorithms↗

PubMed2005

Monte Carlo modeling of the response of NRC's 90Sr/90Y primary beta standard.

The BEAMnrc/EGSnrc Monte Carlo code system is employed to develop a model of the National Research Council of Canada primary standard of absorbed dose to tissue in a beta radiation field, comprising an extrapolation chamber and 90Sr/90Y beta source. We benchmark the model against the measured response of the chamber in terms of absorbed dose to air, for three different experimental setups when irradiated by the 90Sr/90Y source. For the first setup, the chamber cavity depth is fixed at 0.2 cm and the source-to-chamber distance varied between 11 and 60 cm. In the other two cases, the source-to-chamber distance is fixed at 30 cm. In one case the response for different chamber depths is studied, while in the other case the chamber depth is fixed at 0.2 cm as different thicknesses of Mylar are added to the front surface of the extrapolation chamber. The agreement as a function of distance between the calculated and measured responses is within 0.37% for a variation in response of a factor of 29. In the case of dose versus chamber depth, the agreement is within 0.4% for the ISO-recommended nominal depths of 0.025-0.25 cm. Agreement between calculated and measured responses is very good (between 0.02% and 0.2%) for added Mylar foils of thicknesses up to 10.8 mg cm(-2). For larger Mylar thicknesses, deviations of 0.6%-1.2% are observed, which are possibly due to the systematic uncertainties associated with the restricted collisional stopping powers of air or Mylar used in the calculations. We conclude that our simulation model represents the extrapolation chamber and 90Sr/90Y source with adequate accuracy to calculate correction factors for accurate realization of dose rate to tissue at a depth of 7 mg cm(-2) in an ICRU tissue phantom, despite the fact that the uncertainties in the physical characteristics of the source leave some uncertainty in certain calculated quantities.

Computer Simulation↗

PubMed2004

Evidence for using Monte Carlo calculated wall attenuation and scatter correction factors for three styles of graphite-walled ion chamber.

The basic equation for establishing a 60Co air-kerma standard based on a cavity ionization chamber includes a wall correction term that corrects for the attenuation and scatter of photons in the chamber wall. For over a decade, the validity of the wall correction terms determined by extrapolation methods (K(w)K(cep)) has been strongly challenged by Monte Carlo (MC) calculation methods (K(wall)). Using the linear extrapolation method with experimental data, K(w)K(cep) was determined in this study for three different styles of primary-standard-grade graphite ionization chamber: cylindrical, spherical and plane-parallel. For measurements taken with the same 60Co source, the air-kerma rates for these three chambers, determined using extrapolated K(w)K(cep) values, differed by up to 2%. The MC code 'EGSnrc' was used to calculate the values of K(wall) for these three chambers. Use of the calculated K(wall) values gave air-kerma rates that agreed within 0.3%. The accuracy of this code was affirmed by its reliability in modelling the complex structure of the response curve obtained by rotation of the non-rotationally symmetric plane-parallel chamber. These results demonstrate that the linear extrapolation technique leads to errors in the determination of air-kerma.

Algorithms↗

PubMed2004

Large efficiency improvements in BEAMnrc using directional bremsstrahlung splitting.

The introduction into the BEAMnrc code of a new variance reduction technique, called directional bremsstrahlung splitting (DBS), is described. DBS uses a combination of interaction splitting for bremsstrahlung, annihilation, Compton scattering, pair production and photoabsorption, and Russian Roulette to achieve a much better efficiency of photon beam treatment head simulations compared to the splitting techniques already available in BEAMnrc (selective bremsstrahlung splitting, SBS, and uniform bremsstrahlung splitting, UBS). In a simulated 6 MV photon beam (10 x 10 cm2 field) photon fluence efficiency in the beam using DBS is over 8 times higher than with optimized SBS and over 20 times higher than with UBS, with a similar improvement in electron fluence efficiency in the beam. Total dose efficiency in a central-axis depth-dose curve improves by a factor of 6.4 over SBS at all depths in the phantom. The performance of DBS depends on the details of the accelerator being simulated. At higher energies, the relative improvement in efficiency due to DBS decreases somewhat, but is still a factor of 3.5 improvement over SBS for total dose efficiency using DBS in a simulated 18 MV photon beam. Increasing the field size of the simulated 6 MV beam to 40 x 40 cm2 (broad beam) causes the relative efficiency improvement of DBS to decrease by a factor of approximately 1.7 but is still up to 7 times more efficient than with SBS.

Algorithms↗

PubMed2004

Accuracy of the Burns equation for stopping-power ratio as a function of depth and R50.

The accuracy of the Burns et al. equation [Med. Phys. 23, 489-501 (1996)] for the Spencer-Attix water to air stopping-power ratio as a function of depth in a water phantom and electron beam quality in terms of R50 is investigated by comparison to the original data on which this fit was based. It is shown that using this equation provides dose estimates on the central axis in a clinical electron beam that are accurate to within 1% of dose maximum for all 24 clinical beams investigated except very close to the surface in swept beams. In contrast, the error in the dose as a percentage of the local dose is much higher for values of the depth/R50 greater than 1.2.

Algorithms↗

PubMed2004

CSnrc: correlated sampling Monte Carlo calculations using EGSnrc.

CSnrc, a new user-code for the EGSnrc Monte Carlo system is described. This user-code improves the efficiency when calculating ratios of doses from similar geometries. It uses a correlated sampling variance reduction technique. CSnrc is developed from an existing EGSnrc user-code CAVRZnrc and improves upon the correlated sampling algorithm used in an earlier version of the code written for the EGS4 Monte Carlo system. Improvements over the EGS4 version of the algorithm avoid repetition of sections of particle tracks. The new code includes a rectangular phantom geometry not available in other EGSnrc cylindrical codes. Comparison to CAVRZnrc shows gains in efficiency of up to a factor of 64 for a variety of test geometries when computing the ratio of doses to the cavity for two geometries. CSnrc is well suited to in-phantom calculations and is used to calculate the central electrode correction factor Pcel in high-energy photon and electron beams. Current dosimetry protocols base the value of Pcel on earlier Monte Carlo calculations. The current CSnrc calculations achieve 0.02% statistical uncertainties on Pcel, much lower than those previously published. The current values of Pcel compare well with the values used in dosimetry protocols for photon beams. For electrons beams, CSnrc calculations are reported at the reference depth used in recent protocols and show up to a 0.2% correction for a graphite electrode, a correction currently ignored by dosimetry protocols. The calculations show that for a 1 mm diameter aluminum central electrode, the correction factor differs somewhat from the values used in both the IAEA TRS-398 code of practice and the AAPM's TG-51 protocol.

Algorithms↗

PubMed2003

Calculations for plane-parallel ion chambers in 60Co beams using the EGSnrc Monte Carlo code.

The EGSnrc Monte Carlo simulation system is used to obtain, for 10 plane-parallel ionization chambers in 60Co beams, the correction factors Kcomp and Pwall that account for the nonequivalence of the chamber wall material to the buildup cap and the phantom material, respectively. A more robust calculation method has been used compared to that used in previous works. A minor conceptual error related to the axial nonuniformity correction factor, Kan, has been identified and shown to have an effect of about 0.2%. The assumption that Pwall in-phantom is numerically equal to Kcomp calculated for a water buildup cap is shown to be accurate to better than 0.06%, thereby justifying the use of Kcomp calculations which are much more efficient. The effect on the calculated dose to the air in the cavity of the particle production threshold and transport energies used in the simulations is studied. Uncertainties in the calculated correction factors due to uncertainties in the photon and electron cross-section data are studied. They are 0.14% and 0.24%, respectively (1 standard deviation), for Kcomp factors. The uncertainties on Kwall factors are 0.03% from photon cross-section uncertainties and negligible from electron cross-section uncertainties. A comparison with previous EGS4/PRESTA calculations shows that present results are systematically higher by an average of 0.8%, ranging from 0.4% up to 1.4%. The present results are in better agreement with reported experimental values.

Algorithms↗

PubMed2003

Monte Carlo calculated correction factors for primary standards of air kerma.

Many laboratories with cavity chambers as primary standards for air kerma are considering using additional Monte Carlo calculated correction factors, in particular the correction for attenuation and scatter in the walls, Kwall, and possibly the correction for point of measurement, Kan. Standards labs also use Monte Carlo calculated Spencer-Attix stopping-power ratios for graphite to air. The purpose of this article is to investigate the sensitivity of these calculations to their details and to assign uncertainties to the calculated values. We also investigate the correction needed for the Canadian primary standard to account for a polystyrene insulator, Kcomp and find that it is quite large (1.0046 +/- 0.0017). The article shows that the values of correction factors are very robust and insensitive to most details of the calculations except the values of the underlying electron stopping powers which have a significant effect on the stopping-power ratio and on Kcomp. The 1% uncertainties on the photon cross-sections have a negligible effect on these correction factors except for Kcomp. As a result of these investigations, with no change in the stopping power data used, the Canadian primary standard of air kerma in a 60Co beam needs to be increased by 0.54%.

Air↗

PubMed2003

Which accelerator photon beams are "clinic-like" for reference dosimetry purposes?

Previous work has demonstrated that, for photon beam dosimetry, TPR10(20) is not an ideal beam quality specifier for all bremsstrahlung beams, especially for lightly filtered beams in some standards laboratories. This paper addresses the following questions: Is TPR10(20) an adequate beam quality specifier for all modern clinical therapy accelerators? When can nonclinical beams in standards laboratories be used to calibrate ion chambers or measure kQ factors as a function of TPR10(20)? Based on detailed Monte Carlo simulations of Varian, Siemens, Elekta, and GE (Saturn) accelerators one can conclude that TPR10(20) is an adequate beam quality specifier for all these machines in the sense that for a given value of TPR10(20), the value of stopping-power ratios is the same. It is shown that, for the heavily filtered beams used in standards laboratories, TPR10(20) is an adequate beam quality specifier. It is also demonstrated that, for a larger range of bremsstrahlung beams than previously, %dd(10)x is a good beam quality specifier for all clinical beams as well as the lightly and heavily filtered beams in some standards laboratories. A criterion, based on the measured values of TPR10(20) and %dd(10)x for the beam, is proposed for determining whether a nonclinical beam is well specified by TPR10(20). Agreement between calculations for specific accelerators and measured beam quality specifiers is shown to be good, but agreement with published data for a variety of clinical accelerators is not as good. Possible reasons for the discrepancy are discussed.

Energy Transfer↗

PubMed2003

An EGSnrc investigation of cavity theory for ion chambers measuring air kerma.

The EGSnrc system is used to compare the response of an aluminum-walled thimble chamber to that of a graphite-walled thimble chamber for a 60Co beam. When compared to previous experimental results, the EGSnrc values of the ratios of chamber response differ by as much as 0.7% from the experiment. However, it is shown that this difference can be more than accounted for by switching from using the graphite mean excitation energy of 78 eV used in dosimetry protocols to the value of 86.8 eV suggested by more recent stopping-power experiments. This suggests that the uncertainty analysis of Monte Carlo results must be done more carefully, by taking into account uncertainties in the underlying basic data such as the electron and photon cross sections. In comparison to Spencer-Attix cavity theory for a thick-walled ion chamber, the Monte Carlo calculated values of the chamber response differ from the expected ones by 0.15% and 0.01% for the graphite and aluminum chambers, respectively, which are comparable to previously reported values for the Spencer-Attix correction factors. EGSnrc is also used to investigate the effect on the chamber response of thin dag layers on the inside of the aluminum wall. There is good agreement between the calculated and measured changes in chamber response versus the thickness of the dag. The results are compared to the predictions of the Almond-Svensson extension of cavity theory and show that the theory does not correctly predict the chamber response in the presence of thin dag layers. This finding is in agreement with previously reported experimental results. It is demonstrated that the values of alpha, the fraction of ionizations in the gas arising from electrons generated in the dag layer, used in the theory, are not the source of the disagreement.

Algorithms↗

PubMed2002

Multigroup discrete ordinates modeling of 125I 6702 seed dose distributions using a broad energy-group cross section representation.

Our purpose in this work is to demonstrate that the efficiency of dose-rate computations in 125I brachytherapy, using multigroup discrete ordinates radiation transport simulations, can be significantly enhanced using broad energy group cross sections without a loss of accuracy. To this end, the DANTSYS multigroup discrete ordinates neutral particle transport code was used to estimate the absorbed dose-rate distributions around an 125I-model 6702 seed in two-dimensional (2-D) cylindrical R-Z geometry for four different problems spanning the geometries found in clinical practice. First, simulations with a high resolution 210 energy groups library were used to analyze the photon flux spectral distribution throughout this set of problems. These distributions were used to design an energy group structure consisting of three broad groups along with suitable weighting functions from which the three-group cross sections were derived. The accuracy of 2-D DANTSYS dose-rate calculations was benchmarked against parallel Monte Carlo simulations. Ray effects were remedied by using the DANTSYS internal first collision source algorithm. It is demonstrated that the 125I primary photon spectrum leads to inappropriate weighting functions. An accuracy of +/-5% is achieved in the four problem geometries considered using geometry-independent three-group libraries derived from either material-specific weighting functions or a single material-independent weighting function. Agreement between Monte Carlo and the three-group DANTSYS calculations, within three standard Monte Carlo deviations, is observed everywhere except for a limited region along the Z axis of rotational symmetry, where ray effects are difficult to mitigate. The three-group DANTSYS calculations are 10-13 times faster than ones with a 210-group cross section library for 125I dosimetry problems. Compared to 2-D EGS4 Monte Carlo calculations, the 3-group DANTSYS simulations are a 100-fold more efficient. Provided that these efficiency gains can be sustained in three-dimensional geometries, the results suggest that discrete ordinates simulations may have the potential to serve as an efficient and accurate dose-calculation algorithm for low-energy brachytherapy treatment planning.

Brachytherapy↗

PubMed2002

Monte Carlo calculation of nine megavoltage photon beam spectra using the BEAM code.

A recent paper analyzed the sensitivity to various simulation parameters of the Monte Carlo simulations of nine beams from three major manufacturers of commercial medical linear accelerators, ranging in energy from 4-25 MV. In this work the nine models are used: to calculate photon energy spectra and average energy distributions and compare them to those published by Mohan et al. [Med. Phys. 12, 592-597 (1985)]; to separate the spectra into primary and scatter components from the primary collimator, the flattening filter and the adjustable collimators; and to calculate the contaminant-electron fluence spectra and the electron contribution to the depth-dose curves. Notwithstanding the better precision of the calculated spectra, they are similar to those calculated by Mohan et al. The three photon spectra at 6 MV from the machines of three different manufacturers show differences in their shapes as well as in the efficiency of bremsstrahlung production in the corresponding target and filter combinations. The contribution of direct photons to the photon energy fluence in a 10 x 10 field varies between 92% and 97%, where the primary collimator contributes between 0.6% and 3.4% and the flattening filter contributes between 0.6% and 4.5% to the head-scatter energy fluence. The fluence of the contaminant electrons at 100 cm varies between 5 x 10(-9) and 2.4 x 10(-7) cm(-2) per incident electron on target, and the corresponding spectrum for each beam is relatively invariant inside a 10 x 10 cm2 field. On the surface the dose from electron contamination varies between 5.7% and 11% of maximum dose and, at the depth of maximum dose, between 0.16% and 2.5% of maximum dose. The photon component of the percentage depth-dose at 10 cm depth is compared with the general formula provided by AAPM's task group 51 and confirms the claimed accuracy of 2%.

Algorithms↗