RPT 195 Monte Carlo

AAPM REPORT NO.
195
Monte Carlo Reference Data Sets

for Imaging Research
The Report of AAPM Task Group 195

September 2015
DISCLAIMER: This publication is based on sources

and information believed to be reliable, but the
AAPM, the authors, and the editors disclaim any war-
ranty or liability based on or relating to the contents of
this publication.
The AAPM does not endorse any products, manufac-

turers, or suppliers. Nothing in this publication should
be interpreted as implying such endorsement.
© 2015 by American Association of Physicists in Medicine

This page intentionally left blank.
Monte Carlo Reference Data Sets
for Imaging Research
The Report of AAPM Task Group 195
Ioannis Sechopoulos1* (chair), Elsayed S. M. Ali2**, Andreu Badal3, Aldo Badano4,

John M. Boone5, Iacovos S. Kyprianou6, Ernesto Mainegra-Hing7, Michael F. McNitt-Gray8,
Kyle L. McMillan9, D. W. O. Rogers10, Ehsan Samei11, and Adam C. Turner12***
1Emory University, Atlanta, Georgia, USA

2,10Carleton University, Ottawa, Ontario, Canada
3,4,6Food and Drug Administration, Rockville, Maryland, USA
5University of California, Davis, California, USA
7National Research Council Canada, Ottawa, Ontario, Canada
8,9,12University of California, Los Angeles, California, USA
11Duke University, Durham, North Carolina, USA
*Present address: Radboud University Medical Center and the Dutch Reference Centre for Screening
(LRCB)
**Present address: The Ottawa Hospital Cancer Centre
***Present address: Ironwood Cancer and Research Centers
DISCLAIMER: This publication is based on sources and information believed to be reliable,
but the AAPM, the authors, and the publisher disclaim any warranty or liability
based on or relating to the contents of this publication.
The AAPM does not endorse any products, manufacturers, or suppliers. Nothing in this
publication should be interpreted as implying such endorsement.
ISBN: 978-1-936366-38-5
ISSN: 0271-7344
© 2015 by American Association of Physicists in Medicine
All rights reserved
Published by
American Association of Physicists in Medicine

One Physics Ellipse
College Park, MD 20740-3846
THE REPORT OF AAPM TASK GROUP 195:
Monte Carlo Reference Data Sets for Imaging Research
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2. Common Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.1 Monte Carlo Packages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.2 Simulated Material Compositions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.3 X-ray Spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3. Simulation Descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.1 Case 1: Half Value Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.2 Case 2: Radiography and Body Tomosynthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.3 Case 3: Mammography and Breast Tomosynthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.4 Case 4: Computed Tomography with Simple Solids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.5 Case 5: Computed Tomography with a Voxelized Solid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3.6 Case 6: X-ray Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.1 Case 1: Half Value Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.2 Case 2: Radiography and Body Tomosynthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4.3 Case 3: Mammography and Breast Tomosynthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.4 Case 4: Computed Tomography with Simple Solids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.5 Case 5: Computed Tomography with a Voxelized Solid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4.6 Case 6: X-ray Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
5. Discussion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.1 Comparison of Results among Monte Carlo Packages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.2 Lessons Learned . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
5.3 Simulation Times. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
6. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
7. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
APPENDIX: Monte Carlo Package-specific Simulation Parameters . . . . . . . . . . . . 58

A. EGSnrc. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
B. Geant4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
C. MCNP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
D. Penelope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
5
Abstract
The use of Monte Carlo simulations in diagnostic medical imaging research is widespread due to its
flexibility and ability to estimate quantities that are challenging to measure empirically. However, any
new Monte Carlo simulation code needs to be validated before it can be used reliably. The type and
degree of validation required depends on the goals of the research project, but typically, such valida-
tion involves either comparison of simulation results to physical measurements or to previously pub-
lished results obtained with established Monte Carlo codes. The former is complicated due to nuances
of experimental conditions and uncertainty, while the latter is challenging due to typical graphical pre-
sentation and lack of simulation details in previous publications. In addition, entering the field of
Monte Carlo simulations in general involves a steep learning curve. It is not a simple task to learn how
to program and interpret a Monte Carlo simulation, even when using one of the publicly available
code packages. This task group report provides a common reference for benchmarking Monte Carlo
simulations across a range of Monte Carlo codes and simulation scenarios. All simulation conditions
are provided for six different Monte Carlo simulation cases that involve common x-ray-based imaging
research areas. For all cases, the results obtained with four publicly available Monte Carlo software
packages are included in tabular form. In addition to a full description of all simulation conditions and
results, a discussion and comparison of results among the Monte Carlo packages and the lessons
learned during the compilation of these results are included.
This work provides an investigator the necessary information to benchmark his/her Monte Carlo
simulation software against the reference cases included here before performing his/her own novel
research. In addition, an investigator entering the field of Monte Carlo simulations can use these
descriptions and results as a self-teaching tool to ensure that he/she is able to perform a specific simu-
lation correctly. Finally, educators can assign these cases as learning projects as part of course objec-
tives or training programs.
6
1. INTRODUCTION
The use of Monte Carlo methods in diagnostic medical imaging research is an attractive option
because of the relative ease with which many different calculations can be performed. These simula-
tions sometimes span large parameter spaces or obtain estimates of quantities that are not simple to
measure empirically, e.g., absorbed dose or x-ray scatter. The ever-continuing reduction in the cost of
computing power has also helped increase the use of this research methodology.
As with all other types of experimentation, Monte Carlo simulations need to be validated before
their results can be trusted. In many cases, the validation process for a Monte Carlo simulation is not a
simple endeavor. Ideally, Monte Carlo-based computer programs are validated by generating a simula-
tion that replicates an empirical test and then comparing simulation and empirical results. However,
this requires replicating the conditions of a physical experiment to a high level of accuracy and detail
in the computer simulation to minimize differences from the measurement conditions. This level of
replication of a physical experiment can be extremely challenging, even for simple experimental con-
ditions.
One possible step in the validation process of Monte Carlo simulations is the replication of pub-
lished (and previously validated) simulations. However, this approach also presents challenges. In the
first place, a relevant previously published simulation has to exist. Once one is identified, a common
pitfall is the lack of sufficient detail in the description of the simulation, making its replication
extremely challenging. In addition, many times the published results are provided in graphs or sum-
marized form, so an appropriate comparison with the results of the simulation being validated can be
limited and the effort laborious.
To aid researchers in their efforts to validate their Monte Carlo simulations in imaging research or
to assist in the validation process of their use of a general purpose code already available, this report
provides six different reference sets of simulations that include a complete description of the simula-
tion conditions. In addition, the simulation results for all sets—as generated by four commonly used
Monte Carlo packages—are provided in tabulated form along with their statistical uncertainty. These
data sets allow the reader to accurately replicate the geometry, source properties, and scoring. In this
way, the reader can have confidence that his/her simulation is providing accurate results against these
thoroughly vetted reference cases. The agreement amongst the four codes used here indicates the level
of accuracy to be expected as these codes have all been benchmarked against experimental data in dif-
ferent situations.1–10 It is worth to emphasize that as our results were not based on experimental mea-
surements, the validation achieved through this benchmarking is only relative to these calculations
with widely used codes and not directly to measurements. But the concordance between the four
Monte Carlo results provides strong confidence for the value and the validity of this exercise. It is in
that sense that we use the term “validation” throughout the body of this document. In publications and
presentations involving the code, a reference to this report and the obtained level of agreement will be
sufficient to communicate the performance of this component of a validation effort.
Another factor often complicating Monte Carlo simulations is the fact that learning to perform
these simulations involves a steep learning curve. It takes a substantial amount of time and effort to
become comfortable in the programming method for using any of the publicly available Monte Carlo
packages. To help address this, it is hoped that these reference simulation sets can also be used as a
teaching tool, either for a scientist entering the field of Monte Carlo simulations in diagnostic imag-
ing, or for a supervising educator training students or junior researchers in the field. The availability of
the complete descriptions of relevant simulations and their expected results when using different well-
established Monte Carlo codes may be used to check one’s own skills or as an assignment for trainees.
The diagnostic imaging simulation sets included in this report are listed in Table 1. They span a
number of x-ray-based imaging modalities and common quantities scored with Monte Carlo simula-
7
Table 1: List of the simulations sets included in this report
Case Number Modality/Description Quantities Scored

1 Half Value Layer Fluence
(a) Total dose
2 Radiography and Body Tomosynthesis
(b) X-ray primary and scatter photon energy
(a) Total dose and glandular dose
3 Mammography and Breast Tomosynthesis
(b) X-ray primary and scatter photon energy
4 Computed Tomography (simple volumes) Dose
5* Computed Tomography (voxelized volumes) Dose
6 X-ray Production Photon energy fluence and fluence spectra
* Case 5 may also be useful for dosimetry simulations with other imaging modalities (e.g., radiography, body tomosynthesis) that
involve voxelized models of patients.
tions. While this is a limited set that does not encompass all modalities and tests that are commonly
investigated using Monte Carlo methods, the defined conditions provide a meaningful representation
of Monte Carlo-based research. Future reports might be generated to provide additional reference data
sets involving other imaging modalities (e.g., nuclear medicine) or simulation conditions (e.g., detec-
tor simulations).
This report is organized as follows. In Section 2, a description of all the parameters common to the
simulation sets is provided. This includes the Monte Carlo packages used to generate the results and
the material compositions and x-ray spectra used throughout the simulation sets. Section 3 provides a
description of each simulation with all the details necessary to replicate it. Section 4 provides the
results obtained by the task group members for each simulation set. Section 5 includes a discussion on
the common pitfalls encountered in the implementation of the simulations (“lessons learned”) and a
discussion on the differences in the results obtained with the Monte Carlo packages used in this report.
Finally, the Appendix provides the Monte Carlo package-specific parameters used by the task group
members to perform the Monte Carlo simulations used to generate the results included in this report.
2. COMMON PARAMETERS
2.1 Monte Carlo Packages
All simulation conditions described in this report were implemented with four commonly used and
well-established Monte Carlo packages (Table 2). Results from all four implementations of each sim-
ulation are provided in this report. These packages were selected for inclusion in this report due to
their being commonly used by the diagnostic imaging research community, their availability and con-
tinued development and maintenance, and the expertise of the task group members in using them.
Inclusion of these Monte Carlo packages in this report should not be interpreted as any form of
endorsement by the task group or the American Association of Physicists in Medicine (AAPM), and
the exclusion of any other Monte Carlo package should not be interpreted as disapproval by the same.
It should be noted that the results and performance of the four Monte Carlo packages used for this
report should not be necessarily construed to be the best achievable with each package. The results
and performance obtained with these packages reflect the best attempts of the task group members to
match the conditions specified in each case description, with logical choices for each code-specific
parameter. The simulations were not necessarily optimized for minimizing running time or to achieve
the maximum level of accuracy beyond that of typical use. In addition, errors could have been made
during the implementation of the simulations.
8
Table 2: Monte Carlo packages used to generate the results included

in this report, in alphabetical order
Package Photon
Version References Comments
Name Cross Sections
EGSnrc default (XCOM:
Used in all cases except for Case 1, where NIST
Rayleigh, photoelectric and
EGSnrc V4 r2.4.0 11–14 XCOM cross sections were used for all
pair production.
interactions.
RIA: Incoherent scattering.)
Electromagnetic physics Used in all cases except for Case 6, where the
Geant4 9.6 patch 2 15,16
option 4 package Livermore electromagnetic physics were used.
MCNPX 2.7a ENDF/B-VII 17 Used in all cases.
Used in all cases except Case 6. Used with the
Penelope default
2006 penEasy main program adapted for parallel
(XCOM and EPDL)
Penelope 18–20 simulation with MPI library.
Penelope default Used in Case 6, with standard penEasy main
2008
(XCOM and EPDL) program.
2.2 Simulated Material Compositions

The composition of materials used in all simulations is provided in two separate electronic files avail-
able for download in the electronic resources included with this report. One file includes all material
compositions for all cases, except those used in Case #5, which are defined in a separate file.
The densities specified for the elements, as well as the material composition for the simpler com-
pounds, were those provided by the National Institute of Standards and Technology (NIST).21 The
composition of soft tissue used is that provided by the International Commission on Radiation Units
and Measurements (ICRU).22 Due to their common use in breast imaging research, the material compo-
sitions related to mammography and breast tomosynthesis are those provided by Hammerstein et al.23
2.3 X-ray Spectra
The probability distribution functions of the x-ray spectra used as inputs in the simulations included in
this report were obtained using Report 78 of the Institute of Physics and Engineering in Medicine
(IPEM).24 They were defined so as to approximate certain x-ray spectra defined in Publication 61267
of the International Electrotechnical Commission.25 The spectra definitions do not comply with the
IEC 61267 requirement for their homogeneity coefficient, the ratio of the first to the second half value
layer, however all other requirements are met, and this has no effect on the validity of this benchmark-
ing process. Table 3 lists the x-ray spectra used and the parameters specified for the generation of their
distribution functions using the IPEM Report 78 software. The resulting spectra are plotted in Figure
1. The distribution functions themselves are available for download in the electronic resources
included with this report.
The energy bin width for all spectra is 0.5 keV, starting at 0 keV, and the energy of the center of the
bin is listed. During the Monte Carlo simulations, the provided distribution functions should be sam-
pled uniformly within each bin.
9
Table 3: Characteristics of x-ray spectra used for the simulations included in this report. All x-ray spectra are
described with probability distribution functions with bin width of 0.5 keV starting at 0 keV. The mid-bin energy
of the bins is listed in the electronic resource file with the probability distribution functions.
HVL and QVL values are in terms of air kerma determined using planar fluence.
IEC 61267 Name

Characteristics
RQR-8 RQR-9 RQR-M3
Target/Filter Elements W/Al W/Al Mo/Mo
Tube Voltage (kVp) 100 120 30
Anode Angle (deg) 11 11 15
Ripple (%) 0 0 0
Filter Thickness (mm) 2.708 2.861 0.0386
Mean Energy (keV) 50.6 56.4 16.8
Half Value Layer (HVL) (mm Al)* 3.950 5.010 0.3431
Quarter Value Layer (QVL) (mm Al)† 9.840 – 0.7663
* These values are not those reported by IPEM Report 78 software, but were calculated by the task group using the NIST XCOM dataset.21
† Provided for the two spectra used in Case #1 only.
Mo/Mo30kVpXͲRaySpectrum W/AlXͲRaySpectra
0.35 0.06
0.30 W/Al100kVp
0.05
W/Al120kVp
0.25
RelativeFluence
0.04
RelativeFluence
0.20
0.03
0.15
0.02
0.10
0.01
0.05
0.00 0.00
0 5 10 15 20 25 30 0 20 40 60 80 100 120
XͲRayEnergy(keV) XͲRayEnergy(keV)
W/AlXͲRaySpectra
Figure0.06 1. Graphs of the x-ray spectra, normalized to unit fluence under the curve, used in the Monte Carlo simu-
lations included in this report.
W/Al100kVp
3. SIMULATION DESCRIPTIONS
This section provides the complete description of the simulations with all details necessary to replicate
the conditions used in the implementation of each by the task group. All references to materials and x-
ray spectra refer to those defined in Section 2 and as provided in the electronic resources.
10
3.1 Case 1: Half Value Layer

Aim
This case aims to verify the accuracy of input x-ray spec-
trum sampling, basic material attenuation, and half-value
layer calculations.
Geometry
1. Filter is a disk of thickness t mm and diameter 40 mm
(Figure 2). The central axis of the filter coincides with
the +z axis. The superior face of the filter is 100 mm
below the source.
2. Simulations are performed for two thicknesses (t), cor-
responding to the theoretical half-value layer (HVL)
and quarter value layer (QVL), respectively, as given in
Table 4 (also listed in Table 3 with the description of the
x-ray spectra).
Materials
Filter material is aluminum and the rest of the geometry is
filled with air.
Radiation Source
Figure 2. Diagram of the geometry
1. Radiation source is an isotropic x-ray point source at setup for Case #1.
the geometry origin (x=0, y=0, z=0), collimated to a
central circle of 1 mm diameter at the superior face of
the filter.
2. Simulations are performed for monoenergetic photons and for spectra as given in Table 4.
Scoring
1. Scoring is performed in a central circle of 10-mm diameter at the plane z = +1000 mm without the
presence of the filter and with the filters set at the two thicknesses specified in Table 4.
2. By planar energy fluence it is meant ψ = Σi (Ei)/A, where Ei is the energy of the incident x-ray and
A is the area, without consideration for the cosine of the incidence angle.
3. For monoenergetic photons, the quantities scored are as follows:
a. Primary (non-scattered) planar energy fluence (ψp (t)) at the incident photon energy.
b. Scatter (ψs (E, t)) differential planar energy fluence, binned in 0.5 keV bins between zero and
the incident photon energy, to match the bin resolution of the input spectra.
Table 4: Simulation parameters for Case #1
t for Theoretical HVL t for Theoretical QVL

X-ray Energy/Spectrum
(mm Al) (mm Al)
30 keV 2.273 4.546
100 keV 15.11 30.22
Mo/Mo 30 kVp 0.3431 0.7663
W/Al 100 kVp 3.950 9.840
11
4. For x-ray spectra, the quantities scored are the primary (ψp (E, t)) and scatter (ψs (E, t)) differential
planar energy fluences using the same bin definitions as above.
5. All planar energy fluence values are converted to air kerma (AK) off-line, using the air mass
§μ ·
energy absorption coefficients, ¨ en ¸ , available for download in the electronic resources included
© ρ ¹
with this report on a 0.5 keV spacing, as obtained from NIST,21 and the following equation:
§μ ·
AK = E Φ ¨ en ¸ (3.1)
© ρ ¹
where E is the mid-point energy bin of the incident x-ray and Φ is the planar fluence at energy bin E.
Since this conversion is done off-line using the provided coefficients, all energy fluence for each
bin is assumed to be at the mid energy of the bin. As can be seen in the provided electronic file, all
mass energy absorption coefficients for x-ray energies below 3.75 keV are listed as zero, so any
incident photons below this energy may be ignored.
6. Final results are the following four air kerma ratios for each x-ray source energy definition:
a. Primary only, resulting AK ratio after HVL:
¦ AK ( E, t = HVL )
p
R1 = E
(3.2)
¦ AK ( E, t = 0 )
E
p
b. Primary only, resulting AK ratio after QVL:
¦ AK ( E, t = QVL )
p
R2 = E
(3.3)
¦ AK ( E, t = 0 )
E
p
c. Primary and scatter, resulting AK ratio after HVL:
¦ AK ( E, t = HVL ) + ¦ AK ( E, t = HVL )
p s
R3 = E E
(3.4)
¦ AK ( E, t = 0 ) + ¦ AK ( E, t = 0 )
E
p
E
s
d. Primary and scatter, resulting AK ratio after HVL:
¦ AK ( E, t = QVL ) + ¦ AK ( E, t = QVL )
p s
R4 = E E (3.5)
¦ AK ( E, t = 0 ) + ¦ AK ( E, t = 0 )
E
p
E
s
Statistical Uncertainty
The number of simulated x-rays is such that the statistical uncertainty is 1% or lower on the primary
total planar energy fluence in the case of monoenergetic incident photons, and 1% on the primary dif-
ferential planar energy fluence in any given bin in the case of incident spectra.
12
3.2 Case 2: Radiography and Body Tomosynthesis

Aim
This case aims to verify the accuracy of x-ray transport and interaction characteristics in general radi-
ography and whole body tomosynthesis simulations. It scores absorbed dose in the body and the
energy of both x-ray primary and scatter components incident upon the detection plane. Depending on
the application, the dosimetry or the fluence component results may be tested.
Geometry
1. Body is a box of thickness (z-direction) 200 mm, with width (x-direction) and height (y-direction)
390 mm. Its center is located at the x-y coordinates (0, 0) and its entrance surface is 1550 mm
from the x-ray source (figures 3 and 4).
Figure 3. Diagram of the geometry setup for Case #2 with the x-ray source positioned for the radiography acqui-
sition (equivalent to the tomosynthesis 0° angle).
Figure 4. Diagram of the geometry setup for Case #2 with the x-ray source positioned for the tomosynthesis 15°
angle acquisition.
13
2. Scoring plane has 390 mm sides in the x-y directions and is located 50 mm past the body.
Materials
Body material is soft tissue, and the rest of the geometry is filled with air.
Radiation Source
1. Radiation source positions are as follows:
a. For simulation of radiography and a 0° tomosynthesis projection, source is located at the
geometry origin (x=0, y=0, z=0).
b. For simulation of the 15° tomosynthesis projection, source is displaced along the –y direction,
forming an angle of 15° with the scoring plane.
c. These positions apply to both the full field and the point spread function simulations.
2. For validation of dosimetry, the x-ray source is an isotropic point source with the x-ray beam col-
limated electronically for congruence with all four edges of the scoring plane.
3. For validation of x-ray scatter, two types of sources are used:
a. First test (full field): Same as point 2 above for validation of dosimetry.
b. Second test (point-spread function): X-ray point source emits a zero-area beam aimed at the
center of the scoring plane.
4. Simulations are performed for the W/Al 120 kVp spectrum and for monoenergetic photons with
energy 56.4 keV (equivalent to the mean energy of the spectrum).
Figure 5. Diagram showing the locations of the VOIs within the simulated body of Case #2 where the energy
deposition is scored for validation of the dosimetry simulation. VOI 3 is located at the center of the body in all
three directions. VOIs 1, 2, 4, and 5 are at the same z location as VOI 3. VOIs 1, 5, 6, 7, 8, and 9 are at the same x
location as VOI 3.
14
Figure 6. Diagram showing the locations of the ROIs within the scoring plane of Case #2 where the incident pri-
mary and scatter energy is scored for validation of the x-ray scatter simulation with a full-field x-ray beam. ROI 5 is
located at the center of the scoring plane.
Scoring
1. For validation of dosimetry, two sets of scoring results are provided:
a. Total energy deposited in the body per initial photon.
b. Energy deposited in 9 cubic volumes of interest (VOIs) inside the body per initial photon, each
is 27,000 mm3 (30 mm × 30 mm × 30 mm). Five VOIs are located on the central plane parallel
to the scoring plane. Four additional VOIs are located on the central axis perpendicular to the
scoring plane. The locations are shown in Figure 5.
2. When validating x-ray primary and scatter components for the first test using a full-field x-ray
beam, the scoring is as follows:
a. Primary x-ray energy per initial photon, incident on the 7 regions of interest (ROIs), each
30 mm × 30 mm, throughout the scoring plane, according to Figure 6.
b. Scatter x-ray energy per initial photon, incident on the same 7 ROIs.
These ROIs correspond to 30 mm × 30 mm regions with the following image indices:
1: (0, 0)
2: (6, 0)
3: (3, 3)
4: (6, 3)
15
5: (6, 6)
6: (9, 9)
7: (12, 12)
3. When validating x-ray scatter for the second test using a zero-area beam resulting in a point-
spread function, the scoring should be as follows:
a. Incident scatter x-ray energy per initial photon, at 7 ROIs, each 30 mm × 30 mm, throughout
the scoring plane, according to Figure 7.
b. Incident primary x-ray energy per initial photon, at ROI 5, according to Figure 7.
1: (4, 4)
2: (6, 4)
3: (5, 5)
4: (6, 5)
5: (6, 6)
6: (7, 7)
7: (8, 8)
Figure 7. Diagram showing the locations of the ROIs within the scoring plane where the incident primary and scat-
ter energy is scored for validation of the x-ray scatter simulation with a zero-area beam. ROI 5 is located at the
16
4. Secondary results: When possible, a more complete validation of the x-ray scatter simulation can
be achieved by separately scoring the results for:
a. X-rays that underwent only one Compton event.
b. X-rays that underwent only one Rayleigh event.
c. X-rays that underwent more than one scatter event (Compton or Rayleigh).
The number of simulated x-rays is such that the statistical uncertainty is 1% or lower on all scored
quantities at each VOI (dose) and ROI (scatter).
17
3.3 Case 3: Mammography and Breast Tomosynthesis

Aim
This case aims to verify the accuracy of x-ray transport and interaction characteristics in mammogra-
phy and breast tomosynthesis simulations, resulting in the validation of estimates of absorbed dose in
the breast glandular tissue and x-ray scatter incident upon the scoring plane. Depending on the appli-
cation, the dosimetry or the fluence component results may be tested.
Geometry
1. Breast skin is a semi-circular cylinder of thickness (z-direction) 50 mm and radius 100 mm. It is
positioned relative to the scoring plane in the center in the y-direction and with the chest wall side
congruent with the chest wall edge of the scoring plane. This is the typical mammographic posi-
tion for a cranio-caudal (CC) view breast. The entrance surface of the breast skin is located
595 mm below the x-ray source at the mammography position (figures 8, 9, and 10).
2. Breast tissue is another semi-circular cylinder, concentric with the breast skin, with thickness
(z-direction) 46 mm and radius 98 mm. This results in a skin thickness of 2 mm enveloping the
interior breast tissue in all directions except at the chest wall side. The breast tissue is located ver-
tically (z-direction) at the center of the breast skin.
3. The breast compression and breast support plates are rectangular boxes with dimensions in the
(x, y, z) directions of 140 mm × 260 mm × 2 mm. The top plate is located immediately above the
breast skin in the z-direction (bottom surface touching the top of the breast) and the bottom plate
is immediately below the breast skin in the z-direction (top surface touching the bottom of the
breast). The centers of both plates are centered with the scoring plane in the (x, y) directions.
4. The scoring plane has sides of 140 mm × 260 mm in the x-y directions and is located 15 mm
below the bottom of the breast skin.
5. The body is a rectangular box with dimensions in the (x, y, z) directions of 170 mm × 300 mm ×
300 mm, and has its +x surface congruent with the chest wall side of the breast and scoring plane.
Figure 8. Diagram of the geometry setup for Case #3 with the x-ray source positioned for the mammography
acquisition (equivalent to the tomosynthesis 0° angle).
18
It is centered with the center of the breast in the y- and z- directions. This body volume is present
to include any relevant backscatter in the simulations.
Materials
1. The skin of the breast is composed of the material defined as “breast skin.”
Figure 9. Diagram of the front view of the geometry setup for Case #3 with the x-ray source positioned for both
the mammography and the breast tomosynthesis 15° angle acquisitions.
Figure 10. Diagram detailing the dimensions of the simulated breast in Case #3.
19
2. The interior breast tissue is a homogeneous mixture of 80% breast adipose tissue and 20% breast
glandular tissue.
3. The breast compression and support plates are made of PMMA.
4. The body is made of water.
5. The rest of the geometry is filled with air.
Radiation Source
1. Radiation source positions are as follows:
a. For simulation of mammography and 0° tomosynthesis projection, the source is located at the
chest wall edge of the scoring plane in the x-direction, centered with the scoring plane in the
y-direction and at z=0.
b. For simulation of the 15° tomosynthesis projection, the source is rotated 15° about the x-axis,
with the center of rotation at the scoring plane.
c. These positions apply to both the full field and the point spread function simulations.
2. For validation of dosimetry, the x-ray source is an isotropic point source with the x-ray beam col-
limated for congruence with all four edges of the scoring plane.
3. For validation of x-ray scatter, two types of sources are used as follows:
a. First test (full-field): Same as point 2 above for validation of dosimetry.
b. Second test (point-spread function): X-ray point source emits a zero-area beam aimed at the
4. Simulations are performed for the Mo/Mo 30 kVp spectrum and for monoenergetic photons with
Scoring
1. For validation of dosimetry, there are three sets of scoring results:
a. Total energy deposited in the interior breast tissue, excluding the skin, per initial photon.
b. Mean glandular dose per initial photon, using the method described in 1999 by Boone26:
¦ {g ( E inc , G ) × Edep }
MGD =
all hits ,
(3.6)
G × massbreast tissue
where G is the fraction by weight that is glandular tissue and
§μ ·
G ¨ en ¸
© ρ ¹ gland
g ( Einc , G ) = (3.7)
§μ · §μ ·
G ¨ en ¸ + (1 − G ) ¨ en ¸
© ρ ¹ gland © ρ ¹ adip
where
§ μen · 1
¨ ¸= . (3.8)
© ρ
2 3 4 5
¹ a + b × Einc + c × Einc + d × Einc + e × Einc + f × Einc
20
Table 5: Fit coefficient for the mass energy absorption coefficients

for breast adipose and glandular tissue
Fit Coefficients Adipose Tissue Glandular Tissue

a 4.792 × 10−5 1.112 × 10−4
b −9.042 × 10−5 −2.324 × 10−4
c 1.718 × 10−4 2.594 × 10−4
d 2.482 × 10−4 1.296 × 10−4
e 1.090 × 10−5 7.977 × 10−6
f −1.764 × 10−7 −1.489 × 10−7
Einc is the energy, in keV, of the x-ray when the energy deposition occurs, and the fit coeffi-
cients a through f are listed in Table 5. These were obtained by fitting the data provided by
NIST21 using commercial fitting software (TableCurve 2D, Systat Software, San Jose, CA).
These fits were performed for x-rays with an energy range of 1.0 keV to 30.0 keV spaced at
intervals of 0.5 keV.
c. Energy deposited in 7 cubic volumes of interest (VOIs) inside the breast, each 4,000 mm3
(20 mm × 20 mm × 10 mm) per initial photon. Five VOIs are located on the central plane par-
allel to the scoring plane, and two additional VOIs are located directly above and below VOI 3.
The locations are shown in Figure 11.
Figure 11. Diagram showing the locations of the VOIs within the simulated breast in Case #3 where the energy
deposition is scored for validation of the dosimetry simulation. VOI 3 is located at the center of the breast in the y-
and z-directions. VOIs 1, 2, 4, and 5 are at the same z location as VOI 3. VOIs 2, 4, 6, and 7 are at the same y loca-
tion as VOI 3.
21
Figure 12. Diagram showing the locations of the ROIs within the scoring plane in Case #3 where the incident pri-
mary and scatter energy are scored for validation of the x-ray scatter simulation with a full-field x-ray beam.
2. When validating x-ray primary and scatter components for the first test using a full-field x-ray
beam, the scoring is as follows:
a. Primary x-ray energy per initial photon, incident on 7 regions of interest (ROI), each 20 mm
× 20 mm, throughout the scoring plane, according to Figure 12.
b. Scatter x-ray energy per initial photon, incident on the same 7 ROIs.
1: (0, 0)
2: (0, 3)
3: (2, 3)
4: (0, 6)
5: (1, 6)
6: (3, 6)
7: (0, 9)
22
Figure 13. Diagram showing the locations of the ROIs within the scoring plane in Case #3 where the incident pri-
mary and scatter energy are scored for validation of the x-ray scatter simulation with a zero-area beam. ROI 5 is
located at the center of the scoring plane.
3. When validating x-ray primary and scatter results for the second test using a zero-area beam
resulting in a point-spread function, the scoring should be as follows:
a. Scatter x-ray energy per initial photon, incident on 7 ROIs, each 20 mm × 20 mm, throughout
the scoring plane, according to Figure 13.
b. Primary x-ray energy per initial photon, incident on ROI 5, according to Figure 13.
1: (1, 4)
2: (2, 5)
3: (1, 6)
4: (2, 6)
5: (3, 6)
6: (2, 7)
7: (1, 8)
23
4. Secondary results: When possible, a more complete validation of the x-ray primary and scatter
simulation can be achieved by discriminating the results by:
a. X-rays that underwent only one Compton event.
b. X-rays that underwent only one Rayleigh event.
c. X-rays that underwent more than one scatter event (Compton or Rayleigh).
quantities.
24
3.4 Case 4: Computed Tomography with Simple Solids

Aim
This case aims to verify the accuracy of x-ray transport and interaction characteristics in computed
tomography in addition to x-ray source rotation, resulting in the validation of estimates of absorbed
dose in a simple CT phantom.
Geometry
Body phantom is a cylinder of height (z-direction) 3,000 mm (to approximate an infinite cylinder) and
radius 160 mm (Figure 14). The center of the cylinder is placed at the x-ray source isocenter of rota-
tion, which is the origin (x=y=z=0).
Materials
1. The body phantom is made of PMMA.
Radiation Source
1. The source is an isotropic x-ray point source collimated to a fan beam with dimensions, measured
at the center of the body phantom, of width (y-direction) equal to the body diameter (320 mm) and
thickness (z-direction) of t. The values for t are provided in Table 6.
2. The rotation radius of the x-ray source about the isocenter, located at the center of the body phan-
tom, is 600 mm.
3. For the first test with the scoring of dose in four contiguous cylindrical segments, the x-ray source
is located at the 0° position (at coordinates x = −600 mm and y = z = 0), as shown in Figure 14.
Parameter Number Minimum Maximum Increment

Fan beam thickness, t (mm) 2 10 80 70
Projection Angles (deg)
First Test 1 0 0 –
Second Test – Discrete 36 0 350 10
Second Test – Random ∞ 0 360 –
Figure 14. Diagram of the geometry setup for Case #4.
25
Figure 15. Diagram showing the locations of the four cylindrical VOIs within the simulated body phantom in Case
#4 where the energy deposition is scored for the first test of this CT dosimetry simulation. This simulation is only
performed with the x-ray source at the 0° position.
4. For the second test with the scoring of dose in two 10 mm diameter cylinders, two different source
types are simulated:
a. Source rotated 360° about the isocenter in 10° increments, with 36 evenly spaced simulations
performed.
b. Angular position of source is randomly sampled for each x-ray emitted from the continuous
distribution of 360° about the isocenter.
Scoring
1. For the first test, the scoring is the energy deposited in four contiguous cylindrical segments from
a single projection, as shown in Figure 15. Each cylindrical segment is an axial tomographic sec-
tion of the phantom with thickness 10 mm, with centers located at z = 0 mm, z = 10 mm,
z = 20 mm, and z = 30 mm (either in the positive or negative z direction).
26
Figure 16. Diagram showing the locations of the two cylindrical VOIs within the simulated body phantom in Case
#4 where the energy deposition is scored for the second test of this CT dosimetry simulation. One set of simula-
tions is performed with the x-ray source positioned at 36 discrete projection angles, and another simulation is per-
formed with the x-ray source positioned at randomly sampled projection angles about the 360°.
2. For the second test, the scoring is the energy deposited in two 10 mm diameter PMMA cylinders
with height (z-direction) of 100 mm from −50 mm to +50 mm, as shown in Figure 16. One cylin-
der is located at the center of the body phantom and one is located with its center 150 mm from the
center of the body phantom.
quantities at each VOI.
27
3.5 Case 5: Computed Tomography with a Voxelized Solid

Aim
This case aims to verify the accuracy of voxel-based x-ray transport and interaction characteristics in
computed tomography, in addition to x-ray source rotation, resulting in the validation of estimates of
absorbed dose in a complex, voxelized CT phantom.
Even though this simulation uses a relatively thin fan beam, this case may also be useful for verifi-
cation of dosimetry simulations involving voxelized solids in other modalities, such as radiography
and body tomosynthesis. For these, comparison of the results for a single or a limited number of pro-
jection angles may be sufficient.
Geometry
Geometry is exactly the same as that defined for Case #4, but with a voxelized box replacing the cylin-
drical body phantom (Figure 17). The box has dimensions of thickness (x-direction) 320 mm, width
28
Projection Angles (deg) Number Minimum Maximum Increment

Discrete 8 0 345 45
Random ∞ 0 360 –
(y-direction) 500 mm, and height (z-direction) 260 mm, containing 320 × 500 × 260 voxels. This vox-
elized volume contains the description of the torso portion of a human patient. Each voxel is 1.0 mm ×
1.0 mm × 1.0 mm.
Materials
1. The three-dimensional (3D) image with the information for the material content of the voxelized
volume is available for download in the electronic resources included with this report. This refer-
ence case is the XCAT model, courtesy of Ehsan Samei and Paul Segars of Duke University, to
serve as a reference platform for Monte Carlo simulations. Care should be taken to use this vol-
ume with the correct orientation in the Monte Carlo simulation. The voxels in the image contain
values ranging from 0 to 19 that correspond to material definitions also available for download in
the electronic resources included with this report.
Radiation Source
1. The isotropic x-ray point source is collimated to a fan beam with dimensions, measured at the cen-
ter of the voxelized volume, of width (y-direction) equal to the voxelized volume (500 mm) and
thickness (z-direction) of 10 mm.
2. The rotation radius of the x-ray source about the isocenter, located at the center of the body phan-
tom, is 600 mm.
3. The 0° position of the x-ray source is located at coordinates x = −600 mm and y = z = 0, as shown
in Figure 17, and increasing angle projections are in the direction marked in the same figure.
4. Two different source types are simulated (see Table 7):
a. Source rotated 360° about the isocenter in 45° increments, with 8 evenly spaced simulations
performed.
b. Angular position of the source is randomly sampled for each x-ray emitted from the continu-
ous distribution of 360° about the isocenter.
Scoring
The scoring is the energy deposited in all the voxels with values 3 to 19, separated by organ/material.
The number of simulated x rays is such that the statistical uncertainty is 1% or lower on dose scored in
all organ/materials except for the adrenals (voxel value = 12).
29
3.6 Case 6: X-ray Production

Aim
This case aims to verify the accuracy of electron transport and x-ray generation in typical tube targets
in mammography and radiography. The results will depend on the physical models used to simulate
Bremsstrahlung emission and inner shell ionizations, as well as the internal tables on characteristic x-
ray emission energies.
Geometry
1. Target is a slab of height 25 mm, width 80 mm, and thickness 10 mm. The center of the front face
of the slab is at the geometry origin (x=0, y=0, z=0). The slab is tilted at an angle θ relative to the
+z axis (Figure 18).
2. Simulations are performed for two target materials and their corresponding target angles and inci-
dent electron energies, as given in Table 8.
Materials
1. Target material is either molybdenum or tungsten, depending on the incident electron kinetic
energy.
2. The rest of the geometry is filled with a vacuum.
Incident Electron
Modality Target Material Anode Angle (deg)
Kinetic Energy (keV)
Mammography Mo 30 15
Radiography W 100 11
30
Figure 19. Diagram of the scoring areas for Case #6.
Radiation Source
1. Radiation source is an electron beam parallel to the +x direction that uniformly covers a circle of
diameter 3 mm centered at the geometry origin (x=0, y=0, z=0).
2. Simulations are performed for monoenergetic electrons with energies given in Table 8.
Scoring
1. Energy distribution is scored in 1 keV bins for photons crossing the plane z = +100 mm at five
square areas, 10 mm × 10 mm each. The locations of the scoring areas are given in Figure 19.
2. For each square area, the output is the energy distribution and the integral planar energy fluence.
Areas 2, 1, and 3 reflect the heel effect, while areas 4, 1, and 5 reflect symmetry.
The number of simulated source particles is such that the statistical uncertainty is 1% on each bin of
the scored x-ray energy distribution from 7.5 keV and up.
31
4. Results
Detailed tabular results for all six cases simulated with the Monte Carlo packages listed in Table 2 are
available for download in the electronic resources included with this report. To aid in the comparison
of the results between the different Monte Carlo packages included in this report, some tables and
graphs are included here for some specific simulation conditions that are representative of all the
results. Additional graphs are included in the results spreadsheets available for download in the elec-
tronic resources. All uncertainties shown in tables and graphs are due only to the statistical uncertain-
ties from the Monte Carlo simulations and do not include any uncertainties or limitations of the
underlying physics models.
It should be noted that the ranges of results presented in this report are those found from the Monte
Carlo simulations performed as a part of this task group effort. They represent the best attempts of the
task group members. However, they do not necessarily reflect the best possible performance or accu-
racy of each Monte Carlo package. As such, to consider a new simulation validated, the results
obtained do not necessarily have to be within the ranges presented here. It is up to the investigator’s
discretion as to when his/her results should be considered consistent with those found by this task
group, and therefore whether his/her code can be considered validated.
The comparison of simulated results with experimental measurements is still an essential step in
the process of validating a new simulation algorithm. However, the possibility of comparing simula-
tion results among different codes using clearly defined geometries, material composition, and scoring
methods eliminates many of the sources of variability and error found in experimental measurements,
and it provides an insight into the inner workings of the simulation code at a precision level beyond
the accuracy of standard experimental measurements.
All results spreadsheets that are available for download in the electronic resources included with
this report include timing information for the simulations implemented with the four Monte Carlo
packages listed in Table 2. The timing information for each simulation is presented in two forms: (i)
the mean time per history simulated, and (ii) the total simulation time needed to achieve an uncertainty
of 1% for a specific scored quantity. Given the possible use of different variance reduction techniques,
the latter quantity in general provides more complete information regarding simulation times than the
former. The timing results are estimated assuming a sequential execution of the simulation on a single
core of a modern CPU. The simulations can be run serially or in parallel on a cluster of computers and,
therefore, the effective simulation time depends on the amount of computational resources invested in
the execution. Each scored quantity for which the timing to achieve 1% uncertainty is provided is
specified by a color highlight in the appropriate cell in the spreadsheet.
32
4.1 Case 1: Half Value Layer
HalfValueLayerͲ PrimaryOnly
0.502
0.501
0.500
EGSnrc
R1
Geant4
0.499
MCNP
Penelope
0.498
0.497
30keV 100keV 30kVp 100kVp
XͲRayBeam
QuarterValueLayerͲ PrimaryOnly
0.252
0.251
0.250
EGSnrc
R2
0.249
Geant4
MCNP
0.248
Penelope
0.247
0.246
30keV 100keV 30kVp 100kVp
XͲRayBeam
Figure 20. Calculated values of the HVL (top) and QVL (bottom) based on just primary photons in Case #1 for the
four Monte Carlo packages included in Table 2. Note the narrow range of the vertical scale and the presence of the
uncertainty bars due to the statistical uncertainty in each simulation.
33
4.2 Case 2: Radiography and Body Tomosynthesis
TotalBodyEnergyDeposited
33500
EnergyDepositedp.i.p.(eVperphoton)
33000
32500
32000
31500 EGSnrc
Geant4
31000
MCNP
30500 Penelope
30000
29500
0deg 15deg 0deg 15deg
56.4keV W/Al120kVp
ProjectionAngle
Figure 21. Energy deposition in the whole body per initial photon in Case #2 for the four Monte Carlo packages
included in Table 2. Note the narrow range of the vertical scale and the presence of the uncertainty bars due to the
statistical uncertainty in each simulation.
Table 9: Minimum and maximum ratios to the mean values for the energy deposited
in the entire body volume for all Monte Carlo simulations performed for Case #2
with the full-field x-ray beam. Statistical uncertainties are 0.00014% to 0.10%.
Radiography (0 deg) Tomosynthesis (15 deg)

56.4 keV 0.999 – 1.001 1.000 – 1.001
W/Al 120 kVp 0.999 – 1.002 0.999 – 1.001
in each VOI on the body volume for all Monte Carlo simulations performed for Case #2
56.4 keV W/Al 120 kVp

Radiography Tomosynthesis Radiography Tomosynthesis
Body VOI
(0 deg) (15 deg) (0 deg) (15 deg)
1 0.999 – 1.003 0.998 – 1.002 0.997 – 1.002 0.997 – 1.003
2 0.998 – 1.003 0.999 – 1.003 0.997 – 1.003 0.997 – 1.002
3 0.994 – 1.003 0.996 – 1.003 0.990 – 1.007 0.990 – 1.006
4 0.999 – 1.003 0.998 – 1.002 0.997 – 1.003 0.997 – 1.002
5 0.998 – 1.003 0.997 – 1.004 0.997 – 1.002 0.995 – 1.007
6 0.994 – 1.011 0.996 – 1.010 0.995 – 1.012 0.996 – 1.011
7 0.986 – 1.005 0.986 – 1.005 0.982 – 1.010 0.982 – 1.009
8 0.999 – 1.002 0.999 – 1.003 0.996 – 1.005 0.995 – 1.005
9 0.992 – 1.014 0.992 – 1.016 0.992 – 1.020 0.989 – 1.022
34
BodyVOIEnergyDepositedͲ 56.4keV
80
70
60
50
EGSnrc
40
Geant4
30 MCNP
Penelope
20
10
0
1 2 3 4 5 6 7 8 9
BodyVOI
Figure 22. Energy deposition in the VOIs in the body per initial photon for the radiography (0° projection) for the
56.4 keV x-ray energy simulations of Case #2 with the four Monte Carlo packages included in Table 2.
Table 11: Minimum and maximum ratios to the mean values for the primary energy incident
in each ROI at the scoring plane for all Monte Carlo simulations performed for Case #2
with the full field x-ray beam. Statistical uncertainties are 0.004% to 0.61%.

Scoring Plane Radiography Tomosynthesis Radiography Tomosynthesis
ROI (0 deg) (15 deg) (0 deg) (15 deg)
1 0.988 – 1.012 0.999 – 1.001 0.984 – 1.010 0.997 – 1.003
2 0.986 – 1.012 0.998 – 1.002 0.986 – 1.007 0.997 – 1.003
3 0.984 – 1.017 0.988 – 1.014 0.982 – 1.008 0.984 – 1.009
4 0.980 – 1.015 0.987 – 1.014 0.982 – 1.014 0.983 – 1.016
5 0.982 – 1.017 0.985 – 1.015 0.984 – 1.010 0.988 – 1.008
6 0.980 – 1.015 0.987 – 1.014 0.988 – 1.006 0.986 – 1.009
7 0.984 – 1.015 0.986 – 1.015 0.985 – 1.010 0.989 – 1.008
35
PrimaryEnergyͲ 56.4keV
FullFieldBeam
4.5
PrimaryEnergy p.i.p. (eVperphoton) 4.4
4.3
EGSnrc
4.2
Geant4
MCNP
4.1
Penelope
4.0
3.9
1 2 3 4 5 6 7
ScoringPlaneROI
ScatterEnergyͲ 56.4keV
FullFieldBeam
18.0
ScatterEnergyp.i.p.(eVperphoton)
16.0
14.0
12.0
10.0 EGSnrc
8.0 Geant4
6.0 MCNP
Penelope
4.0
2.0
0.0
1 2 3 4 5 6 7
ScoringPlaneROI
Figure 23. Primary component (top) and scatter component (bottom) of energy incident on the scoring plane
ROIs per initial photon with the full field x-ray beam for the radiography (0° projection) 56.4 keV x-ray energy sim-
ulations of Case #2 with the four Monte Carlo packages included in Table 2.
36
PrimaryEnergy
PencilBeam
900
PrimaryEnergy p.i.p. (eVperphoton)
850
800
750
700 EGSnrc
650 Geant4
MCNP
600
Penelope
550
500
56.4keV W/Al120kVp
ScoringPlaneROI
Figure 24. Primary component (both 0° and 15° projections) of photon energy incident on the scoring plane ROI
5 per initial photon with the zero-area beam for both x-ray beams of Case #2 with the four Monte Carlo packages
included in Table 2.
Table 12: Minimum and maximum ratios to the mean values for the scatter energy incident

1 0.995 – 1.010 0.994 – 1.004 0.993 – 1.008 0.993 – 1.006
2 0.995 – 1.008 0.995 – 1.004 0.995 – 1.006 0.995 – 1.007
3 0.997 – 1.004 0.995 – 1.004 0.995 – 1.007 0.994 – 1.007
4 0.998 – 1.005 0.994 – 1.009 0.995 – 1.007 0.994 – 1.006
5 0.996 – 1.007 0.996 – 1.008 0.994 – 1.007 0.994 – 1.008
6 0.995 – 1.009 0.995 – 1.010 0.994 – 1.006 0.993 – 1.006
7 0.993 – 1.012 0.993 – 1.012 0.992 – 1.008 0.993 – 1.012
37
PencilBeam
ScatterEnergyp.i.p.(eVperphoton) 140
120
100
80
EGSnrc
60 Geant4
MCNP
40 Penelope
20
0
1 2 3 4 5 6 7
ScoringPlaneROI
Figure 25. Scatter component of photon energy incident on the scoring plane ROIs per initial photon for the 0°
projection and 56.4 keV x-ray energy with the zero-area beam of Case #2 for the four Monte Carlo packages
included in Table 2.
in ROI 5 at the scoring plane for all Monte Carlo simulations performed for Case #2
with the zero-area x-ray beam. Statistical uncertainties are 0.002% to 0.04%.
Radiography (0 deg) Tomosynthesis (15 deg)

56.4 keV 0.984 – 1.013 0.983 – 1.013
W/Al 120 kVp 0.987 – 1.007 0.986 – 1.007
Table 14: Minimum and maximum ratios to the mean values for the scatter energy incident

1 0.996 – 1.007 0.995 – 1.007 0.996 – 1.007 0.996 – 1.008
2 0.997 – 1.004 0.998 – 1.004 0.996 – 1.008 0.997 – 1.007
3 0.997 – 1.004 0.997 – 1.005 0.995 – 1.006 0.996 – 1.005
4 0.996 – 1.005 0.996 – 1.004 0.997 – 1.005 0.996 – 1.005
5 0.992 – 1.005 0.993 – 1.005 0.999 – 1.001 0.998 – 1.001
6 0.996 – 1.007 0.996 – 1.006 0.996 – 1.006 0.995 – 1.006
7 0.996 – 1.006 0.995 – 1.007 0.997 – 1.008 0.994 – 1.007
38
4.3 Case 3: Mammography and Breast Tomosynthesis
TotalBreastEnergyDeposited
4800
4700
4600
4500
EGSnrc
4400 Geant4
MCNP
4300
Penelope
4200
4100
16.8keV Mo/Mo30kVp
ProjectionAngle
Figure 26. Energy deposition in the breast tissue of Case #3 for the four Monte Carlo packages in Table 2.
in the entire breast volume for all Monte Carlo simulations performed for Case #3
Mammography (0 deg) Tomosynthesis (15 deg)

16.8 keV 0.999 – 1.001 0.997 – 1.002
Mo/Mo 30 kVp 0.999 – 1.001 0.997 – 1.002
Table 16: Minimum and maximum ratios to the mean values for the glandular dose
in the entire breast volume for all Monte Carlo simulations performed for Case #3

16.8 keV 0.999 – 1.002 0.997 – 1.004
Mo/Mo 30 kVp 0.997 – 1.003 0.998 – 1.003
39
BreastVOIEnergyDepositedͲ 16.8keV
20
EnergyDeposited p.i.p. (eVperphoton)

18
16
14
12
EGSnrc
10
Geant4
8
MCNP
6 Penelope
4
0
1 2 3 4 5 6 7
BreastVOI
Figure 27. Energy deposition per initial photon in the VOIs in the breast for the mammography (0° projection) for
the 16.8 keV x-ray energy simulations of Case #3 with the four Monte Carlo packages included in Table 2. Note
that the columns for VOI 7 extend beyond the y-axis scale (to about 56 eV per photon), to enhance visibility of the
results of the other VOIs.
on each VOI in the breast volume for all Monte Carlo simulations performed for Case #3
16.8 keV Mo/Mo 30 kVp

Mammography Tomosynthesis Mammography Tomosynthesis
Breast VOI
(0 deg) (15 deg) (0 deg) (15 deg)
1 0.987 – 1.006 0.983 – 1.009 0.998 – 1.002 0.997 – 1.003
2 0.992 – 1.014 0.984 – 1.009 0.999 – 1.002 0.998 – 1.003
3 0.968 – 1.012 0.983 – 1.008 0.998 – 1.001 0.998 – 1.002
4 0.954 – 1.017 0.983 – 1.008 0.999 – 1.001 0.997 – 1.003
5 0.998 – 1.002 0.982 – 1.012 0.997 – 1.002 0.997 – 1.003
6 0.917 – 1.031 0.985 – 1.009 0.997 – 1.002 0.996 – 1.005
7 0.986 – 1.005 0.980 – 1.009 0.999 – 1.001 0.997 – 1.003
40
Table 18: Minimum and maximum ratios to the mean values for the primary photon energy
incident on each ROI at the scoring plane for all Monte Carlo simulations performed for Case #3

Scoring Plane Mammography Tomosynthesis Mammography Tomosynthesis
1 0.999 – 1.001 0.996 – 1.004 0.999 – 1.001 0.996 – 1.004
2 0.993 – 1.007 0.982 – 1.038 0.992 – 1.007 0.989 – 1.009
3 0.990 – 1.008 0.988 – 1.017 0.992 – 1.009 0.990 – 1.010
4 0.993 – 1.004 0.992 – 1.009 0.990 – 1.009 0.991 – 1.012
5 0.990 – 1.011 0.993 – 1.006 0.993 – 1.008 0.989 – 1.017
6 0.991 – 1.010 0.992 – 1.004 0.995 – 1.006 0.990 – 1.010
7 0.992 – 1.007 0.991 – 1.009 0.991 – 1.009 0.990 – 1.010
Table 19: Minimum and maximum ratios to the mean values for the energy of scattered photons
incident on each ROI at the scoring plane for all Monte Carlo simulations performed for Case #3

1 0.992 – 1.004 0.995 – 1.007 0.989 – 1.010 0.989 – 1.014
2 0.974 – 1.029 0.986 – 1.018 0.981 – 1.022 0.981 – 1.026
3 0.976 – 1.019 0.984 – 1.016 0.984 – 1.015 0.987 – 1.016
4 0.978 – 1.030 0.982 – 1.016 0.980 – 1.027 0.983 – 1.026
5 0.977 – 1.019 0.990 – 1.020 0.985 – 1.014 0.985 – 1.017
6 0.977 – 1.025 0.966 – 1.054 0.987 – 1.010 0.987 – 1.019
7 0.970 – 1.020 0.973 – 1.028 0.980 – 1.021 0.980 – 1.020
41
FullFieldBeam
1.60

1.55
EGSnrc
1.50
Geant4
MCNP
1.45 Penelope
1.40
1 2 3 4 5 6 7
ScoringPlaneROI
FullFieldBeam
1.2
ScatterEnergyp.i.p.(eVperphoton)
1.1
1.0
0.9
EGSnrc
0.8 Geant4
MCNP
0.7 Penelope
0.6
0.5
1 2 3 4 5 6 7
ScoringPlaneROI
Figure 28. Primary component (top) and scattered component (bottom) of photon energy incident per initial pho-
ton on the scoring plane ROIs with the full-field x-ray beam for the mammography (0° projection) 16.8 keV x-ray
energy simulations of Case #3 with the four Monte Carlo packages included in Table 2. Note that the columns for
ROI 1 for both graphs extend beyond the y-axis scale (to about 110 eV per initial photon and 3.4 eV per initial pho-
ton, respectively), to enhance visibility of the results of the other ROIs.
42
PrimaryEnergy
PencilBeam
300

250
200
150 EGSnrc
Geant4
100 MCNP
Penelope
50
0
16.8keV Mo/Mo30kVp
ScoringPlaneROI
Figure 29. Primary only (both 0° and 15° projections) energy incident on the scoring plane ROI 5 per initial pho-
ton with the zero-area beam for both x-ray energies of Case #3 with the four Monte Carlo packages included in
Table 2.
in ROI 5 at the scoring plane for all Monte Carlo simulations performed for Case #3

16.8 keV 0.991 – 1.008 0.991 – 1.008
Mo/Mo 30 kVp 0.993 – 1.007 0.993 – 1.007
43
PencilBeam
ScatterEnergyp.i.p.(eVperphoton) 10
9
8
7
6
EGSnrc
5
Geant4
4
MCNP
3
Penelope
2
1
0
1 2 3 4 5 6 7
ScoringPlaneROI
Figure 30. Scatter component of photon energy incident on the scoring plane ROIs per initial photon for the
mammography (0° projection) 16.8 keV x-ray energy with the zero-area beam of Case #3 for the four Monte Carlo
packages included in Table 2. Note that the columns for ROI 5 extend beyond the y-axis scale (to about 25 eV per
initial photon), to enhance visibility of the results of the other ROIs.
Table 21: Minimum and maximum ratios to the mean values for the scatter component
of the photon energy incident on each ROI at the scoring plane for all Monte Carlo simulations
performed for Case #3 with the zero-area x-ray beam. Statistical uncertainties are 0.01% to 3.77%.

1 0.977 – 1.022 0.981 – 1.038 0.985 – 1.019 0.979 – 1.027
2 0.983 – 1.014 0.984 – 1.013 0.991 – 1.008 0.991 – 1.008
3 0.991 – 1.008 0.985 – 1.014 0.993 – 1.010 0.991 – 1.012
4 0.983 – 1.016 0.985 – 1.016 0.990 – 1.010 0.989 – 1.012
5 0.984 – 1.019 0.983 – 1.020 0.986 – 1.016 0.986 – 1.016
6 0.983 – 1.014 0.985 – 1.013 0.992 – 1.007 0.990 – 1.007
7 0.986 – 1.019 0.989 – 1.024 0.985 – 1.016 0.986 – 1.015
44
4.4 Case 4: Computed Tomography with Simple Solids
FirstTestͲ CylindricalSegment
10mmFanBeam
14000
EnergyDeposited p.i.p. (eVperphoton)
12000
10000
8000
EGSnrc
6000 Geant4
4000 MCNP
Penelope
2000
0
1 2 3 4 1 2 3 4
56.4keV W/Al120kVp
CylindricalSegment
Figure 31. Energy deposited per initial photon in the four cylindrical VOIs of the phantom with the 10 mm thick
fan beam of Case #4 for the four Monte Carlo packages included in Table 2.
in each of the four cylindrical VOIs of the phantom for all Monte Carlo simulations performed
for the first test of Case #4. Statistical uncertainties are 0.02% to 0.79%.

Slice Thickness Slice Thickness Slice Thickness Slice Thickness
Cylindrical VOI
10 mm 80 mm 10 mm 80 mm
1 0.999 – 1.002 0.997 – 1.002 0.998 – 1.002 0.997 – 1.003
2 0.995 – 1.002 0.998 – 1.001 0.994 – 1.003 0.995 – 1.005
3 0.996 – 1.004 0.995 – 1.003 0.997 – 1.003 0.994 – 1.005
4 0.999 – 1.001 0.989 – 1.006 0.995 – 1.004 0.988 – 1.006
in each of the two cylindrical VOIs of the phantom for all Monte Carlo simulations performed
for the second test of Case #4 when using the random continuous x-ray source position.
Statistical uncertainties are 0.27% to 0.81%.

Slice Thickness Slice Thickness Slice Thickness Slice Thickness
Cylindrical VOI
10 mm 80 mm 10 mm 80 mm
Center 0.986 – 1.014 0.983 – 1.017 0.997 – 1.003 0.983 – 1.017
Perimeter 0.986 – 1.014 0.987 – 1.013 0.984 – 1.016 0.979 – 1.021
45
4.5 Case 5: Computed Tomography with a Voxelized Solid
Table 24: Minimum and maximum ratios to the mean values for the energy deposited in each organ/material in the voxelized volume
representing the patient phantom for all Monte Carlo simulations performed for Case #5 when using the random continuous x-ray source position.
These ranges include the results for Geant4 and MCNP, as listed in Table 2. Statistical uncertainties are 0.01% to 4.27%.
Voxel Material
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Soft Heart Lung Liver Gallbladder Spleen Stomach Large Pancreas Adrenal Thyroid Thymus Small Esophagus Skin Breast Cortical
Tissue Intestine Intestine Bone
0.994 − 0.999 − 0.998 − 0.998 − 0.994 − 0.997 − 0.998 − 0.999 − 0.999 − 0.961 − 0.988 − 0.989 − 0.998 − 0.996 − 0.992 − 0.995 − 0.996 −
56.4 keV
1.006 1.001 1.002 1.002 1.006 1.003 1.002 1.001 1.001 1.039 1.002 1.011 1.002 1.004 1.008 1.005 1.004
W/AI 120 0.994 − 0.999 − 0.997 − 0.998 − 0.999 − 0.996 − 0.998 − 0.998 − 0.997 − 0.985 − 1.000 − 0.985 − 0.995 − 0.999 − 0.991 − 0.995 − 0.996 −
kVp 1.006 1.001 1.003 1.002 1.001 1.004 1.002 1.002 1.003 1.015 1.000 1.015 1.005 1.001 1.009 1.005 1.004
Table 25: Minimum and maximum ratios to the mean values for the energy deposited in each organ/material in the voxelized volume
representing the patient phantom for all Monte Carlo simulations performed for Case #5 for the simulation with discrete x-ray source positions
with the monoenergetic 56.4 keV source. Statistical uncertainties are 0.002% to 5.07%.
46
Voxel Material
Projection 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Angle Soft Heart Lung Liver Gallbladder Spleen Stomach Large Pancreas Adrenal Thyroid Thymus Small Esophagus Skin Breast Cortica
Tissue Intestine Intestine l Bone
0.999 − 0.996 − 0.997 − 0.995 − 0.994 − 0.996 − 0.994 − 0.993 − 0.994 − 0.957 − 0.988 − 0.985 − 0.992 − 0.994 − 0.999 − 0.992 − 0.998 −
0
1.002 1.008 1.006 1.010 1.014 1.009 1.012 1.016 1.012 1.030 1.005 1.006 1.017 1.009 1.001 1.018 1.003
0.999 − 0.996 − 0.997 − 0.995 − 0.988 − 0.996 − 0.995 − 0.985 − 0.994 − 0.963 − 0.993 − 0.986 − 0.995 − 0.997 − 0.999 − 0.989 − 0.998 −
45
1.002 1.008 1.005 1.010 1.023 1.010 1.010 1.018 1.011 1.016 1.006 1.008 1.011 1.005 1.001 1.024 1.004
0.999 − 0.993 − 0.995 − 0.995 − 0.994 − 0.991 − 0.992 − 0.995 − 0.993 − 0.974 − 0.984 − 0.993 − 0.997 − 0.994 − 0.999 − 0.995 − 0.996 −
90
1.002 1.014 1.008 1.012 1.012 1.015 1.017 1.012 1.017 1.049 1.014 1.004 1.005 1.012 1.001 1.009 1.009
0.994 − 0.997 − 0.998 − 0.997 − 0.996 − 0.993 − 0.998 − 0.998 − 0.999 − 0.974 − 0.996 − 0.984 − 0.993 − 0.998 − 0.994 − 0.990 − 0.995 −
135
1.003 1.005 1.002 1.005 1.006 1.006 1.004 1.004 1.002 1.037 1.004 1.006 1.005 1.002 1.003 1.004 1.002
0.998 − 0.993 − 0.996 − 0.994 − 0.985 − 0.994 − 0.995 − 0.992 − 0.998 − 0.993 − 0.995 − 0.995 − 0.992 − 0.995 − 0.999 − 0.984 − 1.000 −
180
1.002 1.015 1.008 1.011 1.027 1.012 1.011 1.017 1.004 1.005 1.007 1.009 1.016 1.011 1.001 1.031 1.001
0.995 − 0.997 − 0.998 − 0.996 − 0.985 − 0.997 − 0.997 − 0.998 − 0.995 − 0.988 − 0.996 − 0.989 − 0.997 − 0.996 − 0.996 − 0.994 − 0.996 −
225
1.003 1.003 1.002 1.007 1.013 1.005 1.003 1.002 1.004 1.006 1.003 1.010 1.004 1.006 1.002 1.014 1.002
0.998 − 0.994 − 0.996 − 0.995 − 0.978 − 0.995 − 0.995 − 0.991 − 0.993 − 0.950 − 0.990 − 0.998 − 0.994 − 0.995 − 0.999 − 0.989 − 0.996 −
270
1.003 1.013 1.008 1.015 1.040 1.013 1.013 1.021 1.014 1.062 1.025 1.002 1.008 1.010 1.002 1.026 1.009
0.995 − 0.998 - 0.998 − 0.997 − 0.996 − 0.997 − 0.995 − 0.992 − 0.992 − 0.987 − 0.991 − 0.985 − 0.994 − 0.998 − 0.995 − 0.994 − 0.996 −
315
1.002 1.003 1.002 1.005 1.005 1.009 1.007 1.008 1.007 1.013 1.010 1.006 1.007 1.001 1.003 1.012 1.002
0° ProjectionAngle,56.4keV
1.E+5
EnergyDeposited p.i.p. (eVperphoton) 1.E+4
1.E+3
1.E+2
EGSnrc
1.E+1
Geant4
1.E+0 MCNP
Penelope
1.EͲ1
11Pancreas
19CorticalBone
14Thymus
3Softtissue
4Heart
9Stomach
18Breast
13Thyroid
15SmallIntestine
6Liver
10LargeIntestine
8Spleen
17Skin
12Adrenal
16Esophagus
5Lung
7Gallbladder
VoxelValue/Organ
Figure 33. Energy deposited per initial photon in the voxelized phantom with the 10 mm thick fan beam and the
56.4 keV x-ray energy at the 0° projection angle only of Case #5, for the four Monte Carlo packages included in
Table 2.
47
SecondTestͲ CenterCylinder,DiscreteAngles
10mmFanBeam,56.4keV
12.4
12.3
12.2
EGSnrc
12.1
Geant4
12.0 MCNP
Penelope
11.9
11.8
0 50 100 150 200 250 300 350
ProjectionAngle(deg)
SecondTestͲ PerimeterCylinder,DiscreteAngles
10mmFanBeam,56.4keV
120
100
80
EGSnrc
60
Geant4
40 MCNP
Penelope
20
0
0 50 100 150 200 250 300 350
ProjectionAngle (deg)
Figure 32. Energy deposited per initial photon in center (top) and perimeter (bottom) cylindrical VOI of the phan-
tom with the 10 mm thick fan beam and the 56.4 keV x-ray energy of Case #4 for the four Monte Carlo packages
included in Table 2. Note the very narrow range of the vertical scale of the top graph, showing the similarity among
all results.
48
4.6 Case 6: X-ray Production
TotalEnergyFluenceͲ MoTarget30keV
3.7EͲ04
EnergyFluencep.i.e.[eV/mm2perelectron]
3.5EͲ04
3.3EͲ04
EGSnrc
3.1EͲ04
Geant4
MCNP
2.9EͲ04
Penelope
2.7EͲ04
2.5EͲ04
1 2 3 4 5
ScoringPlaneROI
TotalEnergyFluenceͲ WTarget100keV
4.0EͲ03
EnergyFluencep.i.e. [eV/mm2perelectron]
3.8EͲ03
3.6EͲ03
3.4EͲ03
3.2EͲ03
EGSnrc
3.0EͲ03
Geant4
2.8EͲ03
MCNP
2.6EͲ03 Penelope
2.4EͲ03
2.2EͲ03
2.0EͲ03
1 2 3 4 5
ScoringPlaneROI
Figure 34. Total energy fluence per initial electron at the scoring plane for the Mo target (top) and W target (bot-
tom) simulations of x-ray production of Case #6 for the four Monte Carlo packages included in Table 2.
Table 26: Minimum and maximum ratios to the mean values for the planar energy fluence
at the scoring plane for all Monte Carlo simulations performed for Case #6.
Statistical uncertainties are 0.05% to 0.47%.
ROI 1 ROI 2 ROI 3 ROI 4 ROI 5

Mo target, 30 keV 0.924 – 1.101 0.918 – 1.155 0.925 – 1.067 0.929 – 1.099 0.920 – 1.107
W target, 100 keV 0.990 – 1.017 0.980 – 1.039 0.994 – 1.007 0.995 – 1.011 0.992 – 1.020
49
MoTarget30keV
5.0EͲ09
Fluence p.i.e. [ph/mm2perelectron]

4.0EͲ09
3.0EͲ09 EGSnrc
Geant4
MCNP
2.0EͲ09
Penelope
IPEMRpt.78
1.0EͲ09 Boone
0.0E+00
0 5 10 15 20 25 30
MidͲBinEnergy[keV]
WTarget100keV
2.5EͲ09
Fluence p.i.e. [ph/mm2perelectron]
2.0EͲ09
1.5EͲ09 EGSnrc
Geant4
1.0EͲ09 MCNP
Penelope
IPEMRpt.78
5.0EͲ10
0.0E+00
0 20 40 60 80 100
MidͲBinEnergy[keV]
Figure 35. Photon fluence spectrum per initial electron in ROI 1 for 30 keV electrons on the Mo target (top) and
100 keV electrons on the W target of Case 6 for the four Monte Carlo packages included in Table 2 and as esti-
mated by two spectrum models.24,27 The two spectra models include some inherent filtration, so differences at low
x-ray energies are not relevant.
50
Table 27: Relative energy fluence at ROI 1, 2, and 3 (setting the central ROI [#1] as 100%)
showing the heel effect for the four Monte Carlo packages included in Table 2
Mo Target ROI EGSnrc Geant4 MCNP Penelope

3 108.3% 105.8% 102.4% 106.8%
1 100.0% 100.0% 100.0% 100.0%
2 84.9% 87.0% 91.9% 86.3%
Uncertainty 0.14% 0.28% 0.66% 0.35%
W Target ROI EGSnrc Geant4 MCNP Penelope
3 110.8% 109.5% 108.4% 109.4%
1 100.0% 100.0% 100.0% 100.0%
2 76.9% 77.6% 79.3% 76.6%
Uncertainty 0.13% 0.07% 0.51% 0.31%
5. Discussion
5.1 Comparison of Results among Monte Carlo Packages
In general, the implementations of all six cases with the four Monte Carlo packages included in this
report resulted in estimates with very good concordance, with most of the results being within the sta-
tistical uncertainty of the simulations. For those differences that exceeded the statistical uncertainty,
many differences are within ~5% and most within 10% of the mean of the results. Some, if not all, of
the differences beyond statistical uncertainty could be due to small differences in interpretation of the
simulation conditions or even errors in implementation of the cases by the task group members. Every
attempt was made to minimize this possibility. In some of the cases, the differences were also clearly
produced by the underlying differences in the physical models used by the multiple codes, such as
small differences in the energy of the emitted characteristic x-ray or the modeling of Compton scatter-
ing events.
In the electronic version of the results available for download, the results of each implementation
are provided to allow the readers to analyze the case-by-case concordance for any of the results in
detail. Here we present a summary of the findings:
5.1.1 Case 1: Half Value Layer
The estimated primary and total planar fluence values for all simulations among all four Monte Carlo
implementations matched well, and the resulting air kerma ratios were all very close to the expected
values of 0.500 and 0.250 for the HVL and QVL, respectively.
The thicknesses of the aluminum filter included in the case description were calculated analyti-
cally using the NIST XCOM dataset.21 Therefore, it was expected that the Monte Carlo packages that
use the same cross-section databases, such as GEANT4 and MCNP, would yield air kerma ratios very
close to 0.5 and 0.25 for the half and quarter value layer thicknesses, respectively. Since EGSnrc
(default) and Penelope incoherent scattering model is based on the Relativistic Impulse Approxima-
tion (RIA)13,28−30, their Compton cross sections are different from those in the NIST XCOM dataset.
As a consequence, their half and quarter value layers differ from the expected values by a small but
statistically significant difference (see Table 28). In order to reproduce the expected values, EGSnrc
calculations were run in a NIST-like mode by using XCOM Compton cross sections (see Section A in
the Appendix)31. In this mode, the values obtained are much closer to the expected values.
51
Table 28: Comparison of EGSnrc and Penelope values of the HVL and QVL
for 100 keV photons calculated taking into account the primary photons only
(R1 and R2). Statistical uncertainties of these results are 0.006% and 0.000035%
for Penelope and EGSnrc values, respectively.
HVL (primaries only) QVL (primaries only)

R1 R2
EGSnrc (NIST) 0.49973 0.24973
EGSnrc (default) 0.49792 0.24793
Penelope 0.49781 0.24721
5.1.2 Case 2: Radiography and Body Tomosynthesis

The dose results for all codes and both projection angles yielded very similar results, with almost all
estimates well within 2% of each other. The x-ray scatter study also shows excellent agreement among
all codes at the 2% level. The only notable deviation from this excellent agreement was found to be for
the single Compton scatter estimate of the ROI #5 in the zero area beam simulations (this result can be
found in the electronic version of the results). For this particular ROI, the value obtained with
Penelope is 6% and 8% larger than the value from EGSnrc and Geant4, respectively. This ROI is the
one toward which the beam was directed, so scoring this quantity involves the estimation of very nar-
row-angle Compton scattering. It is possible that there is a larger variation in the physics models
among the packages for this physical interaction, but the cause for these larger discrepancies for this
ROI requires further investigation and is beyond the scope of this report.
As in Case 1, the use of different Compton scattering models by the different Monte Carlo codes
has an impact on the resulting primary x-ray energy incident on the detector. Although the differences
shown in Figures 23 and 24 are small, (2.4% to 3.6% for the full field and 2% to 3% for the pencil
beam), it can be seen in Figure 36 that results obtained with EGSnrc using a NIST-like physics model
FullFieldBeam
4.5
4.4
PrimaryEnergy(eVperphoton)
4.3
4.2
EGSnrc
4.1 Geant4
MCNP
4.0
Penelope
3.9 EGSnrcNIST
3.8
3.7
1 2 3 4 5 6 7
ScoringPlaneROI
Figure 36. Effect of two different photon physics models on the primary energy incident per incident photon on
the scoring plane with the full-field 56.4 keV x-ray beam for the radiography (0° projection) simulation of Case #2.
52
are closer to Geant4 and MCNP results, while EGSnrc results using default parameters for Compton
scattering are closer to Penelope results.
The mono-energetic primary photon beam reaching the scoring plane is very sensitive to differ-
ences in the photon cross sections since it is directly proportional to e−μx, where μ is the total photon
cross section and x is the photon path length in the attenuating material. For this 20 cm thick phantom,
a 1% change in μ translates into a greater than 4% change in e−μx.
5.1.3 Case 3: Mammography and Breast Tomosynthesis
In general, the dose results for all codes, projection angles, monoenergetic and spectral simulations,
and both beam types yielded very similar results, with almost all estimates well within 2% of each
other. Again, the estimation of narrow-angle Compton scattering appears to show a larger variation in
the physics models (see variation in ROI #5 for the zero area beam). At the lower energies used in this
simulation, the full-field estimation of Compton scatter also yields larger differences than the other
estimates. Therefore, the Compton scatter models show greater variability at both narrow scatter
angles and low x-ray energies.
5.1.4 Case 4: Computed Tomography with Simple Solids
Practically all results from all four Monte Carlo packages were found to be within the statistical uncer-
tainty.
5.1.5 Case 5: Computed Tomography with Voxelized Solid
Except for the adrenals, for which the dose estimates had larger statistical uncertainty, the results for
the energy deposition for practically all other organs for both the random and discrete projection
angles yielded very similar results, within 1% to 2% for all codes.
It is worth noting that the estimation of the radiation dose deposited inside the voxels of computa-
tional human phantoms is one of the most relevant and common types of simulation performed in
medical physics research. The results obtained highlight the precision and reliability of dosimetric
studies performed with Monte Carlo simulation codes. As expected, the estimation of dose deposition
inside millimeter-sized voxels is not very sensitive to the differences in the physics used by different
codes.
5.1.6 Case 6: X-Ray Production
This simulation case is more sensitive to subtle differences in physics models compared to the other
simulations because it is sensitive to electron transport details. Thus it is to be expected that the com-
parison among the different Monte Carlo simulation packages would be the most challenging. Differ-
ences among the predicted energy-discriminated fluence were found to be approximately ±10% to
20%, except for the very low-energy bins below approximately 3 keV for the Mo spectrum and 12 keV
for the W spectrum, respectively. At these very low energies, where the electron interaction physics
models are known to have larger uncertainties, the discrepancies between codes were considerably
larger. At these energies, however, diagnostic imaging conditions normally result in these x-rays being
fully attenuated by added x-ray tube filtration, so these discrepancies should not impact realistic imag-
ing research. It should be noted that the results with MCNP resulted in characteristic emission peaks
for the Mo target that were substantially higher than those predicted by the other codes.
While the total energy fluence estimates for the W target spanned a range of approximately ±1%
to ±2%, with only one difference being 4%, the Mo target spectrum simulations demonstrated much
larger differences, mostly due to the substantially higher estimates in characteristic emission by the
MCNP simulations. These calculations are sensitive to the electron impact ionization cross sections
used.
53
5.2 Lessons learned

The exercise performed for this task group report, specifically the design and implementation of sev-
eral simulation test cases with a number of different Monte Carlo simulation packages, resulted in a
number of problems that were identified and corrected along the way. These problems and inconsis-
tencies in interpretation now provide information that we hope will help readers when implementing
their own Monte Carlo simulations.
Chief among the problems that could be encountered is that the specifications of the simulations
may not be followed accurately. Preliminary results of some of the cases included in this report
resulted in large differences among the codes which had to be identified and corrected until the task
group was confident that remaining discrepancies were present either due to statistical uncertainties or
real differences in either the physics tables of each package or some other aspect inside each code. An
example of actual differences in the codes is the small difference in the energy of some characteristic
emission x-rays which resulted in associated peaks corresponding to different energy bins depending
on the Monte Carlo package used. Therefore, utmost care should be taken in implementing these sim-
ulations with the correct input parameters in terms of geometry, materials, source, and scoring. Errors
observed during this exercise, and that can be easily repeated, include the following:
1. Incorrect source type (e.g., isotropic vs. non-isotropic point source). For all simulations except
Case 6 and the point spread function sources (cases 2 and 3), the source should be isotropic.
2. Incorrect scoring metric (e.g., fluence vs. planar fluence vs. energy fluence). For comparison to
the present data, all fluence scoring should be planar fluence, i.e., the cosine factor does not apply.
3. Incorrect scoring units (e.g., total x-rays incident in an ROI vs. x-rays/mm2). Care should be taken
in using the same units as those shown in the provided results spreadsheets.
4. Incorrect binning (e.g., does the energy of a bin represent the mid-level, minimum, or maximum
of a bin? All spectral information, i.e., the x-ray spectra provided for the simulations and the
results for cases 1 and 6 specify the mid-level of the energy bin.
5. Incorrect material composition and density definition. Care should be taken when defining the
materials.
6. Direction of travel in rotating/translating sources. The directions in which rotations or translations
are positive are marked in the figures.
7. X-ray source translation vs. rotation. Note that to more accurately reflect the way the correspond-
ing imaging systems function, the simulation of body tomosynthesis calls for the x-ray source to
translate parallel to the scoring plane to an angle of 15°, while the simulation of breast tomosyn-
thesis involves the source rotating about a point at the scoring plane to an angle of 15°.
8. Incorrect normalization. All results normalized “per photon history” are normalized to only the
acceptable histories that are emitted toward the area specified (in most cases, the entire scoring
plane). The results should not be normalized to all histories potentially emitted in all directions
covering the 4π sphere. The EGSnrc collimated source gives quantities normalized to the incident
real fluence, not the number of histories. To convert to a quantity normalized per number of parti-
cles, one has to calculate the solid angle subtended by the point source and the scoring plane.
9. Excluding the surrounding air was found to have a non-negligible effect on the results in cases 1
and 4.
One issue that can introduce differences in the results is the setting of the energy or range thresh-
old/cut. This is a very important parameter in many simulations, and it can have a substantial impact
54
on the results in many types of Monte Carlo simulations. Among the cases included in this report, this
is especially true for Case #6, the simulation of the generation of x-rays.
An interesting issue was observed with MCNP for Case #4. For all simulations for this case, a void
collimator (material of the collimator set as a void region where photons have no importance) was
used to properly collimate the isotropic x-ray point source. When the collimator and point source were
rotated in 10° increments about the phantom, large differences between results for MCNP and the
other codes were observed between 60−80° and 280−300°. These differences were eliminated when
the phantom inserts, not the collimator and point source, were rotated in 10° increments. It appears
that the rotation of a rigid shape (like the collimator assembly) in an environment with some finite spa-
tial resolution may result in unexpected behavior when photons interact with the edges of the rotated,
rigid structure. Although no clear explanation is available to account for this observation, care should
be taken, at least with MCNP, when using a rotated collimator assembly.
Finally, care should be taken when incorporating variance reduction techniques in the simulations.
During the early stages of obtaining the results for this report, a bug in the implementation of the vari-
ance reduction technique of “particle splitting” in Geant4 version 9.5 and 9.5 patch 1 was found. Due
to this bug, the reduced weight given to each split photon was not inherited by any subsequently cre-
ated secondary particle by these photons, resulting in biased results. This was only identified during the
comparison of the results of the different packages performed during this work, but could have gone
unnoticed in regular research work. The bug has been corrected in later versions of Geant4, including
the one used in the final simulations of all cases included in this report (version 9.6 patch 2).
5.3 Simulation Times
It was not the goal of this task group to compare the efficiency of the Monte Carlo packages tested, but
rather to provide consistent results among them, with the priority being that the simulation conditions
be as similar to each other as possible. Therefore, the use of the most efficient methods to perform
these simulations, including the use of variance reduction techniques, varied considerably among the
groups implementing the simulations. For example, the Penelope simulations were not run with any
variance reduction techniques, and the simulations included the tracking of electrons, although their
inclusion is not expected to affect the results. Although it was not always the case, some of the other
implementations did include variance reduction and did not include tracking of electrons. This
resulted in the Penelope simulation times being larger in general than those for the other codes. The
specific simulation parameters used in the implementation of all cases for each Monte Carlo package
are provided in the Appendix.
In the interest of disseminating as much information as possible, however, we have included in
each results spreadsheet some timing information. Specifically, the spreadsheets include the average
time taken per single history simulated and the amount of time that would be taken by a single, mod-
ern CPU to achieve a 1% statistical uncertainty for a single or a number of scored quantities per case.
Which scored quantity is the one that the timing to 1% precision is referred to is marked in the spread-
sheets by a color highlight. These timing data should only serve as a rough guide, since they were not
produced on the same computer and, as mentioned above, the efficiency of the simulations was not a
priority. In addition, estimating the timing to 1% precision from simulation times that resulted in
scores with orders of magnitude higher precision may result in a substantial underestimation of
required simulation time. However, useful observations can be made. In particular, in various situa-
tions, variance reduction techniques have the ability to very substantially improve the efficiency of
calculations. For example, a factor of 2,000x improvement was obtained in the efficiency of the
EGSnrc calculations in case #6 (x-ray spectra) by using variance reduction techniques. While using
variance reduction methods that are part of general purpose code systems, one must take great care
55
when implementing them in a new code and verify their accurate implementation by comparison to
runs with the technique turned off.
6. Conclusion
This task group developed and implemented six simulations of typical experiments performed for
diagnostic imaging research with x-rays. These simulations were implemented with four different
well-known and publicly available Monte Carlo packages, and the results were compared. All details
of the inputs and the outputs to the simulations are provided in this report and in associated electronic
files.
It is the intent of this task group to provide researchers who are using Monte Carlo simulations
with a standardized methodology to validate their simulations before embarking on research. We
believe that providing these data will eliminate the need to (a) search the literature for potential com-
parable simulations, (b) make informed decisions of the many details that are normally not included in
scientific articles, (c) obtain numerical data from published graphs, and (d) contact publication authors
in hopes of gaining more information on the relevant inputs and results of simulations performed, in
some cases, many years ago.
Of course, each research investigation has different requirements, and therefore the Monte Carlo
simulations needed may include higher levels of complexity compared to those presented in this
report. For example, the simulations described in Case 3 (mammography and breast tomosynthesis)
may be completely sufficient for investigations of dose in a homogeneous breast. On the other hand,
Case 5 (computed tomography dosimetry with a voxelized patient model) may not be sufficient to esti-
mate dose to patients from helical CT scans using automatic exposure control techniques or with the
presence of a complex bow-tie filter. If higher levels of complexity are required, then it is up to the
investigator to determine which additional validation steps would be needed (e.g., comparisons with
physical measurements). Even in these situations in which the cases presented here are not sufficient
for complete validation, comparisons to the reference cases described in this report are strongly rec-
ommended. Therefore, this task group recommends the following:
1. All Monte Carlo methods used for medical imaging research for publication in scientific journals
be analyzed by comparing results obtained with the proposed methods to results of the relevant
case(s) included in this report.
2. The analyses be included in published and presented works, with a reference to the specific case
used, and a description of the relevant details of the case, including x-ray spectra, imaged objects,
tallies performed, and statistics of the Monte Carlo results, as well as the relative percentage dif-
ference found between the results obtained and the corresponding mean of the results in this
report. If the results obtained are outside the ranges of results presented in this report, the relative
percentage deviations from these ranges also need to be provided and justified.
3. This task group suggests the following language to describe the comparison: “The Monte Carlo
methods used in this work were tested by performing case(s) XX-YY of the American Association
of Physicists in Medicine Task Group Report 195 (ref). The simulations were performed using …
(provide details of simulation as listed in point 2 above). The results of our simulations agree to
within X% of the mean results published by TG-195 and they (either) fall within the range of
results published by TG-195 (or) are Y% from the lower/upper limit of the range of results pub-
lished by TG-195. The statistical uncertainty obtained for these simulations was Z%.” A table list-
ing the detailed results of this analysis with appropriate uncertainty estimates is preferred.
4. In published and presented works, the Monte Carlo code and methods being used be described in
detail including software and hardware, total computation time, and any optional parameters being
56
used (e.g., variance reduction techniques, energy cutoff values). Any relevant modification of the
standard code release (such us a new scoring routine) should be described with enough detail that
it could be reproduced by independent investigators.
5. In published works, reference to this report should be made by including a citation to the paper
describing this task group report in the journal Medical Physics.
This task group report can also be useful as an educational tool for trainees and scientists gaining
experience in the field of Monte Carlo simulations, either as a self-teaching tool to ensure that he/she
is able to correctly perform a specific simulation, or as assigned learning projects as part of course
objectives or training programs.
Of course, many imaging modalities and potential simulation types are not included in this report.
This first undertaking had to be limited to a manageable number of cases to be completed in a realistic
time frame. In the future, similar reports that include simulations of other types of measurements or
other imaging modalities may be published by the AAPM.
7. Acknowledgments
The task group would like to thank Samir Abboud, Cecilia Marini Bettolo, Bahaa Ghammraoui,
Sebastien Incerti, Vladimir Ivantchenko, Alfonso Mantero, Luciano Pandola, Joseph Perl, and Yakun
Zhang for their contributions to this work. We thank Eric Jablonowski for his contributions to the fig-
ures. This work was supported in part by the U.S. Department of Energy under contract number DE-
AC02-76SF00515.
57
APPENDIX:
Monte Carlo Package-specific Simulation Parameters
The following is a listing of the specific simulation parameters for the different Monte Carlo packages
used to obtain the results included in this report. Depending on the requirements of each package, the
number and type of parameters can and do vary widely.
A. EGSnrc
1. All cases were simulated with EGSnrc April 2013 release V4-r2-4-0.
2. Photons were followed down to 1 keV, and no electron transport was included in the simulations
except in Case 6, where electrons were followed down to 1 keV. Default Monte Carlo transport
parameters were used for all cases, except for Case 1 where NIST XCOM Compton photon cross
sections were used to better match expected HVL and QVL values.
For case 6, x-ray production, the default electron impact ionization cross section by Kawrakow
was used, although using the Penelope cross section option32 increased the total photon fluence by
6% or 7% for the Mo target case, and the fluorescent peaks by much more.
3. The user-code cavity was used in HVL calculation mode for Case 1, and tutor7pp was used for
cases 2 and 4. For the mammography case (Case 3), the user-code egs_mammo was derived from
tutor7pp to allow for glandular dose and solid angle calculations. Ausgab scoring object
egs_dose_scoring was developed to score dose in cases 4 and 5. A C++ version of ctcreate, ctcre-
ate_raw, was developed that allows creation of egsphant files from a raw binary image. The
BEAMnrc user-code was used to model x-ray production from W and Mo targets (Case 6). For
this case, spectral distribution, particle, and energy fluences were obtained in two different ways,
one using beamdp and a phase-space file, and another using cavity and a newly developed ausgab
scoring object egs_fluence_scoring. Both approaches produced identical results within statistical
uncertainties.
4. Variance reduction techniques were used to increase the efficiency of the simulations. Photon
emission from point sources into a specific scoring field (cases 2, 3, 4, and 5) was simulated using
a collimated source algorithm, uniformly sampling directions to the scoring field and adjusting the
statistical weight accordingly to account for the spatial dependency of the point source photon flu-
ence (EGS_CollimatedSource from the EGSnrc C++ class library). Directional Bremsstrahlung
Splitting (DBS)33,34 and Bremsstrahlung Cross Section Enhancement (BCSE)3 increased the effi-
ciency of x-ray spectra calculations (Case 6) with BEAMnrc. Forced detection was turned on in
cavity for HVL calculations, allowing a very efficient ray-tracing from the source to the scoring
field.
5. The collimated source used with EGSnrc (particles are only emitted in the scoring field direction),
combined with the forced detection scoring scheme, gives the right answer even with one history
for a mono-energetic source. This is equivalent to an analytical ray‐tracing, and because the solid
angle subtended by the scoring field and the source is very small, the maximum variation that one
can get is 0.0005% and 0.0008% with two histories for the 0 cm and thickest filter, respectively.
As a consequence, the statistical uncertainty reported for several million histories is extremely
small, and the 1% statistical uncertainty is just unreachable using this approach. On the other
hand, if one were to use an isotropic source, then the time needed for a 1% statistical uncertainty
in the primary fluence would be 865 sec.
58
6. Statistical uncertainties were calculated using the “history by history” methods detailed by Wal-
ters et al.35
B. Geant4
1. All cases simulated with the Geant4 toolkit for Monte Carlo simulations used version 9.6 patch 2
of the package.
2. All cases were simulated with the electromagnetic physics option 4 package, except for Case 6,
for which the Livermore electromagnetic physics package was used.
3. For all cases except case #6, the cut values used for all particles were the default ones used by this
physics package. For Case #6, the cut values were set to 1.0 µm and a low edge of 1.0 keV.
4. In all cases, electron transport was simulated, although this was not needed except in Case #6.
5. In all cases but Case #6, no variance reduction methods were used. For Case #6, the uniform
Bremsstrahlung splitting algorithm provided in this version of Geant4 was used. It was confirmed
that the results were not biased by this method.
6. No pre-programmed sensitive detectors included in the toolkit were used for these simulations; for
each case, the required sensitive detector was created with custom-made code.
7. For Case #5, the voxelized navigation was performed with the Nested Parameterisation option.
8. All elements were built using the provided NIST material manager, with the compounds defined
from these elements as specified in this report.
9. Statistical uncertainties for all scoring was performed by implementing the method described by
Sempau et al.36
C. MCNP
1. All cases were simulated using MCNPX v2.7a.17
2. Electron transport was simulated only in Case #6.
3. Within all the simulations, the detailed photon transport mode with a low-energy cutoff of 1 keV
was used. The detailed physics treatment included coherent scattering and accounted for fluores-
cent photons after photoelectric absorption. Form factors and Compton profiles were used to
account for electron binding effects. The incoherent, coherent, and photoelectric cross-section
data are based on ENDF/B-IV.37
4. For all cases involving dose calculations, simulation physics options were set so that the photon
transport mode does not explicitly create photoelectrons, but instead assumes all secondary elec-
trons deposit their energy at the photon interaction site, which is reasonable given the incident
photon energy distribution. This assumption satisfies charged particle equilibrium and allows
absorbed dose to be approximated as collision kerma, which was calculated in each volume of
interest by tallying the photon energy fluence and multiplying by the material-specific and energy-
dependent mass energy-absorption coefficient.21
5. Modifications were made to the standard MCNPX source code in order to randomly sample the
angular position of the x-ray source in cases 4 and 5.
D. Penelope
1. All simulations were performed using PENELOPE 2006 (except for case 6, which used
PENELOPE 2008) and the penEasy main program.18−20 The variance reduction methods included
in penEasy were not used, and no particular effort was made to optimize the simulation speed to
avoid compromising the accuracy of these reference results. The user-defined parameters that con-
59
trol the accuracy of the PENELOPE electron and photon transport routines were conservatively
chosen to produce reliable results, even if the simulation speed was reduced. The cut-off energies
for simulating transport of electrons and photons were separately determined for each simulation
depending on the detector system used. For example, in Case #5, the absorption energies were
quite large, 30 keV for electrons and 5 keV for photons, because the scoring volumes were big
compared to the residual range of particles at these energies. In the cases that required an accurate
simulation of low-energy radiation, such as the half value layer estimation and the x-ray produc-
tion, a detailed simulation down to 1 keV for electrons and 0.5 keV for photons was performed.
2. The specific version of the penEasy main program used in this study was penEasy_Imaging, an
extension of penEasy for the study of medical imaging applications (free software is available at
http://code.google.com/p/peneasy-imaging/). To reduce the time required to complete the simula-
tions, the main program was modified to use the MPI libraries to execute the simulations in paral-
lel in a cluster of computers. The most time-consuming cases were executed in parallel in more
than 200 CPU cores to obtain the results in a reasonable amount of time. It was confirmed that the
results of the parallel executions and sequential runs with the same random number seeds were
identical.
3. Some of the simulated cases required small modifications of the main program to score the param-
eters of interest in the way specified in the case description. In particular, the fluence estimation
tally defined in penEasy was not used to estimate the fluence because of the use of different defi-
nitions for this magnitude. In most cases, Python scripts had to be developed to post-process the
simulation output and obtain the desired data in the appropriate format. It was in the post-process-
ing and data collection step where the majority of user errors were made while preparing the pre-
liminary results for this report. The availability of benchmark data from other groups was essential
to detect seemingly trivial errors, such as inadequate normalization, wrong units, geometric errors
in the location of the volumes of interest, or the use of incorrect material files. Some of these
errors resulted in orders of magnitude errors in the results that could have been overlooked in nor-
mal situations, but were easily fixed after the benchmark.
60
References
1
J. R. Mercier, D. T. Kopp, W. D. McDavid, S. B. Dove, J. L. Lancaster, and D. M. Tucker, “Modifica-
tion and benchmarking of MCNP for low-energy tungsten spectra.” Med. Phys. 27, 2680–2687
(2000).
2
G. A. P. Cirrone, et al. “Precision validation of Geant4 electromagnetic physics.” In Nuclear Science
Symposium Conference Record, 2003 IEEE, (2003), vol. 48, p. 482–485.
3
E. S. Ali and D. W. Rogers. “Efficiency improvements of x-ray simulations in EGSnrc user-codes
using bremsstrahlung cross-section enhancement (BCSE).” Med. Phys. 34, 2143–2154 (2007).
4
C. M. Salgado, C. C. Conti, and P. H. B. Becker. “Determination of HPGe detector response using
MCNP5 for 20–150 keV X-rays.” Appl. Radiat. Isot. 64, 700–705 (2006).
5
S. España, J. L. Herraiz, E. Vicente, J. J. Vaquero, M. Desco, and J. M. Udias, “PeneloPET, a Monte
Carlo PET simulation tool based on PENELOPE: features and validation.” Phys. Med. Biol. 54,
1723 (2009).
6
A. Loehr, J. Durst, T. Michel, G. Anton, and P. Geithner. “Comparison of recent experimental data
with Monte Carlo tools such as RoSi, Geant4 and Penelope.” In Nuclear Science Symposium
Conference Record (NSS/MIC), 2009 IEEE, (2009), p. 3904–3908.
7
M. Batic, G. Hoff, M. G. Pia, and P. Saracco. “Photon Elastic Scattering Simulation: Validation and
Improvements to Geant4.” IEEE Trans. Nucl. Sci. 59, 1636–1664 (2012).
8
K. Sangroh, T. Y. Terry, Y. Fang-Fang, and J. C. Indrin. “Spiral computed tomography phase-space
source model in the BEAMnrc/EGSnrc Monte Carlo system: implementation and validation.”
Phys. Med. Biol. 58, 2609 (2013).
9
G. Weidenspointner, Batic, x, M., S. Hauf, G. Hoff, M. Kuster, M. G. Pia, and P. Saracco, “Validation
of Compton scattering Monte Carlo simulation models.” In Nuclear Science Symposium and
Medical Imaging Conference (NSS/MIC), 2013 IEEE, (2013), p. 1–5.
10
R. Schmidt, J. Wulff, and K. Zink. “GMctdospp: Description and validation of a CT dose calcula-
tion system.” Med. Phys. 42, 4260–4270 (2015).
11
D. W. O. Rogers, B. A. Faddegon, G. X. Ding, C. M. Ma, J. We, and T. R. Mackie. “BEAM: a
Monte Carlo code to simulate radiotherapy treatment units.” Med. Phys. 22, 503–524 (1995).
12
I. Kawrakow, “Accurate condensed history Monte Carlo simulation of electron transport. I. EGSnrc,
the new EGS4 version.” Med. Phys. 27, 485–498 (2000).
13
I. Kawrakow, E. Mainegra-Hing, D. W. O. Rogers, F. Tessier, and B. R. B. Walters. “The EGSnrc
Code System: Monte Carlo simulation of electron and photon transport, NRC Technical Report
PIRS-701 v4-2-3-2.” (Ottawa, 2011).
14
D. W. O. Rogers, B. Walters, and I. Kawrakow. “BEAMnrc Users Manual, NRC Technical Report
PIRS-509(A) rev. L.” (Ottawa, 2011).
15
S. Agostinelli, et al. “Geant4 – A simulation toolkit.” Nucl. Instrum. Meth. A 506, 250–303 (2003).
16
J. Allison, et al. “Geant4 Developments and Applications.” IEEE Trans. Nucl. Sci. 53, 270–278
(2006).
17
D. B. Pelowitz, J. S. Hendricks, J. W. Durkee, M. R. James, M. L. Fensin, G. W. McKinney, S. G.
Mashnik, and L. S. Waters. “MCNPX 2.7.A Extensions – Report LA-UR-08-07182.” (Los Ala-
mos, New Mexico, 2008).
18
J. Sempau, E. Acosta, J. Baro, J. M. Fernández-Varea, and F. Salvat. “An algorithm for Monte Carlo
simulation of coupled electron-photon transport.” Nuclear Instruments and Methods in Physics
Research Section B: Beam Interactions with Materials and Atoms 132, 377–390 (1997).
19
F. Salvat, J. M. Fernández-Varea, and J. Sempau. “PENELOPE-2011: A Code System for Monte
Carlo Simulation of Electron and Photon Transport.” (OECD-NEA, Issy-les-Moulineaux,
France, 2011). (Available in PDF at http://www.oecd-nea.org/lists/penelope.html.)
61
20
J. Sempau, A. Badal, and L. Brualla. “A PENELOPE-based system for the automated Monte Carlo
simulation of clinacs and voxelized geometries-application to far-from-axis fields.” Med. Phys.
38, 5887–5895 (2011).
21
J. H. Hubbell and S. M. Seltzer. “Tables of X-Ray Mass Attenuation Coefficients and Mass Energy-
Absorption Coefficients from 1 keV to 20 MeV for Elements Z = 1 to 92 and 48 Additional Sub-
stances of Dosimetric Interest.” July 2004 [accessed December 2007]. (Available from
http://physics.nist.gov/PhysRefData/XrayMassCoef/cover.html.)
22
International Commission on Radiation Units and Measurements. ICRU Report 46: Photon, elec-
tron, proton, and neutron interaction data for body tissues. International Commission on Radia-
tion Units and Measurements, Bethesda, MD, 1992.
23
G. R. Hammerstein, D. W. Miller, D. R. White, M. E. Masterson, H. Q. Woodard, and J. S. Laugh-
lin. “Absorbed radiation dose in mammography.” Radiology 130, 485–491 (1979).
24
K. Cranley, B. J. Gilmore, G. W. A. Fogarty, and L. Desponds. “Catalogue of diagnostic x-ray spec-
tra and other data.” Report No. 78. Institute of Physics and Engineering in Medicine, York,
1997.
25
International Electrotechnical Commission. “IEC 61267 – Medical diagnostic x-ray equipment—
radiation conditions for use in the determination of characteristics.” Geneva, Switzerland, 2005.
26
J. M. Boone. “Glandular breast dose for monoenergetic and high-energy X-ray beams: Monte Carlo
assessment.” Radiology 213, 23–37 (1999).
27
J. M. Boone, T. R. Fewell, and R. J. Jennings. “Molybdenum, rhodium, and tungsten anode spectral
models using interpolating polynomials with application to mammography.” Med. Phys. 24,
1863–1874 (1997).
28
R. Ribberfors. “Relationship of the relativistic Compton cross section to the momentum distribution
of bound electron states.” Physical Review B 12, 2067–2074 (1975).
29
F. Biggs, L. B. Mendelsohn, and J. B. Mann. “Hartree-Fock Compton profiles for the elements.”
Atomic Data and Nuclear Data Tables 16, 201–309 (1975).
30
D. Brusa, G. Stutz, J. A. Riveros, J. M. Fernández-Varea, and F. Salvat. “Fast sampling algorithm for
the simulation of photon Compton scattering.” Nucl. Instrum. Meth. A 379, 167–175 (1996).
31
E. S. M. Ali. “Appendix C: Refining NIST XCOM photon cross sections in EGSnrc.” In Accurate
unfolding of photon spectra of clinical linear accelerators using transmission measurements.
Ph.D. thesis, Carleton University, Ottawa, 2012.
32
D. Bote and F. Salvat. “Calculations of inner-shell ionization by electron impact with the distorted-
wave and plane-wave Born approximations.” Physical Review A 77, 042701 (2008).
33
I. Kawrakow, D. W. Rogers, and B. R. Walters. “Large efficiency improvements in BEAMnrc using
directional bremsstrahlung splitting.” Med. Phys. 31, 2883–2898 (2004).
34
E. Mainegra-Hing and I. Kawrakow. “Efficient x-ray tube simulations.” Med. Phys. 33, 2683–2690
(2006).
35
B. R. B. Walters, I. Kawrakow, and D. W. O. Rogers. “History by history statistical estimators in the
BEAM code system.” Med. Phys. 29, 2745–2752 (2002).
36
J. Sempau, A. Sánchez-Reyes, F. Salvat, H. O. b. Tahar, S. B. Jiang, and J. M. Fernández-Varea.
“Monte Carlo simulation of electron beams from an accelerator head using PENELOPE.” Phys.
Med. Biol. 46, 1163 (2001).
37
J. H. Hubbell, W. J. Veigele, E. A. Briggs, R. T. Brown, D. T. Cromer, and R. J. Howerton. “Atomic
form factors, incoherent scattering functions, and photon scattering cross sections.” Journal of
Physical and Chemical Reference Data 4, 471–538 (1975).
62

RPT 195 Monte Carlo

Uploaded by

Copyright:

Available Formats

RPT 195 Monte Carlo

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

RPT 195 Monte Carlo

Uploaded by

Copyright:

Available Formats

AAPM REPORT NO.

Monte Carlo Reference Data Sets

The Report of AAPM Task Group 195

DISCLAIMER: This publication is based on sources

The AAPM does not endorse any products, manufac-

© 2015 by American Association of Physicists in Medicine

The Report of AAPM Task Group 195

Ioannis Sechopoulos1* (chair), Elsayed S. M. Ali2**, Andreu Badal3, Aldo Badano4,

1Emory University, Atlanta, Georgia, USA

© 2015 by American Association of Physicists in Medicine

All rights reserved

American Association of Physicists in Medicine

APPENDIX: Monte Carlo Package-specific Simulation Parameters . . . . . . . . . . . . 58

Table 1: List of the simulations sets included in this report

Case Number Modality/Description Quantities Scored

Table 2: Monte Carlo packages used to generate the results included

2.2 Simulated Material Compositions

IEC 61267 Name

3.1 Case 1: Half Value Layer

Table 4: Simulation parameters for Case #1

t for Theoretical HVL t for Theoretical QVL

b. Primary only, resulting AK ratio after QVL:

c. Primary and scatter, resulting AK ratio after HVL:

d. Primary and scatter, resulting AK ratio after HVL:

3.2 Case 2: Radiography and Body Tomosynthesis

3.3 Case 3: Mammography and Breast Tomosynthesis

where G is the fraction by weight that is glandular tissue and

Table 5: Fit coefficient for the mass energy absorption coefficients

Fit Coefficients Adipose Tissue Glandular Tissue

3.4 Case 4: Computed Tomography with Simple Solids

Table 6: Simulation parameters for Case #4

Parameter Number Minimum Maximum Increment

Figure 14. Diagram of the geometry setup for Case #4.

3.5 Case 5: Computed Tomography with a Voxelized Solid

Figure 17. Diagram of the geometry setup for Case #5.

Table 7: Simulation parameters for Case #5

Projection Angles (deg) Number Minimum Maximum Increment

3.6 Case 6: X-ray Production

Table 8: Simulation parameters for Case #6

Figure 18. Diagram of the geometry setup for Case #6.

Figure 19. Diagram of the scoring areas for Case #6.

4.1 Case 1: Half Value Layer

4.2 Case 2: Radiography and Body Tomosynthesis

Radiography (0 deg) Tomosynthesis (15 deg)

56.4 keV W/Al 120 kVp

56.4 keV W/Al 120 kVp

PrimaryEnergy p.i.p. (eVperphoton) 4.4

56.4 keV W/Al 120 kVp

Radiography (0 deg) Tomosynthesis (15 deg)

56.4 keV W/Al 120 kVp

4.3 Case 3: Mammography and Breast Tomosynthesis

Mammography (0 deg) Tomosynthesis (15 deg)

Mammography (0 deg) Tomosynthesis (15 deg)

EnergyDeposited p.i.p. (eVperphoton)

16.8 keV Mo/Mo 30 kVp

16.8 keV Mo/Mo 30 kVp

16.8 keV Mo/Mo 30 kVp

PrimaryEnergy p.i.p. (eVperphoton)

PrimaryEnergy p.i.p. (eVperphoton)

Mammography (0 deg) Tomosynthesis (15 deg)