Note: This copy is for your personal, non-commercial use only.

To order presentation-ready copies for

䡲 MEDICAL PHYSICS distribution to your colleagues or clients, use the Radiology Reprints form at the end of this article.

Radiation Dose to the Fetus for

Pregnant Patients Undergoing
Multidetector CT Imaging: Monte
ORIGINAL RESEARCH

Carlo Simulations Estimating Fetal Dose

for a Range of Gestational Age and
Patient Size1
Erin Angel, MS
Purpose: To use Monte Carlo simulations of a current-technology
Clinton V. Wellnitz, MD
multidetector computed tomographic (CT) scanner to in-
Mitchell M. Goodsitt, PhD
vestigate fetal radiation dose resulting from an abdominal
Nazanin Yaghmai, MD and pelvic examination for a range of actual patient anato-
John J. DeMarco, PhD mies that include variation in gestational age and maternal
Christopher H. Cagnon, PhD size.
James W. Sayre, DrPH
Dianna D. Cody, PhD Materials and Institutional review board approval was obtained for this
Donna M. Stevens, MS Methods: HIPAA-compliant retrospective study. Twenty-four mod-
Andrew N. Primak, PhD els of maternal and fetal anatomy were created from image
Cynthia H. McCollough, PhD data from pregnant patients who had previously under-
gone clinically indicated CT examination. Gestational age
Michael F. McNitt-Gray, PhD
ranged from less than 5 weeks to 36 weeks. Simulated
helical scans of the abdominal and pelvic region were
performed, and a normalized dose (in milligrays per 100
mAs) was calculated for each fetus. Stepwise multiple
linear regression was performed to analyze the correlation
of dose with gestational age and anatomic measurements
of maternal size and fetal location. Results were compared
with several existing fetal dose estimation methods.

Results: Normalized fetal dose estimates from the Monte Carlo

simulations ranged from 7.3 to 14.3 mGy/100 mAs, with
an average of 10.8 mGy/100 mAs. Previous methods
yielded values of 10 –14 mGy/100 mAs. The correlation
between gestational age and fetal dose was not signifi-
1
From the Departments of Radiology and Radiation On- cant (P ⫽ .543). Normalized fetal dose decreased lin-
cology, David A. Geffen School of Medicine, University of early with increasing patient perimeter (R2 ⫽ 0.681, P ⬍
California at Los Angeles, 924 Westwood Blvd, Suite 650, .001), and a two-factor model with patient perimeter
Los Angeles, CA 90024 (E.A., N.Y., J.J.D., C.H.C., J.W.S.,
and fetal depth demonstrated a strong correlation with
M.F.M.); Department of Diagnostic Radiology, Mayo Clinic,
Scottsdale, Ariz (C.V.W.); Department of Radiology, Uni- fetal dose (R2 ⫽ 0.799, P ⬍ .002).
versity of Michigan Hospitals, Ann Arbor, Mich (M.M.G.);
Division of Diagnostic Imaging, University of Texas M. D. Conclusion: A method for the estimation of fetal dose from models of
Anderson Cancer Center, Houston, Tex (D.D.C., D.M.S.); actual patient anatomy that represented a range of gesta-
and Department of Radiology, Mayo Clinic College of tional age and patient size was developed. Fetal dose cor-
Medicine, Rochester, Minn (A.N.P., C.H.M.). From the
related with maternal perimeter and varied more than
2006 RSNA Annual Meeting. Received September 24,
2007; revision requested December 10; revision received
previously recognized. This correlation improves when
January 28, 2008; accepted April 7; final version ac- maternal size and fetal depth are combined.
cepted April 21. Supported by grants R01EB004898 and
T32EB002101 from the National Institute of Biomedical 娀 RSNA, 2008
Imaging and Bioengineering. Address correspondence
to E.A. (e-mail: [email protected] ).

姝 RSNA, 2008

220 radiology.rsnajnls.org ▪ Radiology: Volume 249: Number 1—October 2008

MEDICAL PHYSICS: Monte Carlo Simulation in Pregnant Patient Models Angel et al

D
iagnostic computed tomographic modified to represent a newly pregnant edge, this model has only been used for
(CT) imaging is sometimes neces- patient and a patient who was 3 months the estimation of radiation dose in ex-
sary in a pregnant patient. When a pregnant. periments that involved radiation
radiologist needs to decide if the diagnos- Other approaches to estimate fetal sources other than CT.
tic benefits will outweigh the risks of radi- dose are based on Monte Carlo simula- In summary, existing methods for the
ation, it is important to have a reasonably tions of geometric patient models. In estimation of fetal dose for pregnant pa-
accurate estimate of the radiation dose these methods, dose to the uterus is used tients undergoing CT examinations as-
that the conceptus (fetus or embryo) will as an estimate of dose to the fetus at an sume early term pregnancy in a single-
receive. Furthermore, in cases in which early gestational age. One such method is size patient model with an average, non-
pregnancy is discovered during or after CT the ImPACT CTDosimetry dose calcula- varying maternal anatomy. These dose
examination, the patient and/or physician tor (CTDosimetry.xls, version 0.99⫻; Im- estimates do not take into account natural
may request an estimate of the radiation PACT, London, England) (4), which is variations, such as fetal presentation and
dose received by the conceptus. For the based on Monte Carlo simulations per- gestational age. Differences in these at-
remainder of this article, the term fetus will formed by the National Radiological Pro- tributes can cause overestimation or un-
be used to refer to either an embryo or a tection Board (5) with the use of a geo- derestimation of up to 100% for fetal dose
fetus and will therefore be used to describe metric Medical Internal Radiation Dose at radiologic examinations (12).
a conceptus at any gestational age. (MIRD) phantom model (6). Another To overcome these limitations, simu-
It is not known definitively how similar method is the CT-Expo software lations should model specific scanner and
much radiation dose a fetus receives (version 1.5.1; Medizinische Hochschule, maternal and fetal characteristics. The
during CT examination, because this Hannover, Germany) (7,8), which esti- purpose of this study was to use Monte
cannot be measured directly. Some mates organ dose based on simulations Carlo simulations with a current-technol-
methods to estimate fetal dose exist, but performed by Zankl et al at the German ogy multidetector CT scanner to estimate
these estimates are limited by their sim- National Research Center with the Eva fetal dose at abdominal and pelvic CT ex-
plifying assumptions. Existing fetal dose geometric phantom model to represent a amination with models of pregnant pa-
estimation methods are based on phan- standard-size female patient (9,10). tients that represented a range of gesta-
tom measurements and/or geometric The MIRD and Eva phantoms con- tional ages and maternal sizes.
phantom simulation methods. sist of geometric shapes representing
The method used to estimate fetal human organs for a standard-size pa-
dose developed by Felmlee et al (1) is tient model. Although these geometric Materials and Methods
based on anthropomorphic phantom phantoms do not represent pregnant
Patient Images
measurements and measured CT dose in- patient anatomy, ImPACT CTDosim-
Twenty-four voxelized models of mater-
dex (2). More recently, Hurwitz et al (3) etry and CT-Expo software can be used
nal and fetal anatomy were created on
estimated fetal dose by using physical to estimate radiation dose to the uterus,
measurements from internal dosimeters which may be used to approximate fetal
in an anthropomorphic phantom that was dose during the first 8 weeks of gesta- Published online
tion. A patient-based voxelized model of 10.1148/radiol.2491071665
a patient who is 30 weeks pregnant does Radiology 2008; 249:220 –227
Advances in Knowledge exist (11), but, to the best of our knowl-
䡲 Radiation dose to the fetus at CT Abbreviation:
MIRD ⫽ Medical Internal Radiation Dose
imaging in pregnant patients was
estimated with examples of actual Implications for Patient Care
Author contributions:
pregnant patient anatomy that 䡲 For a typical abdominal and pelvic Guarantors of integrity of entire study, E.A., D.D.C.,
represented a range of gestational CT examination, the average fetal M.F.M.; study concepts/study design or data acquisition
or data analysis/interpretation, all authors; manuscript
ages. dose is approximately 24 mGy
drafting or manuscript revision for important intellectual
䡲 Radiation dose to the fetus corre- with a range of 16 –31 mGy, de- content, all authors; manuscript final version approval, all
lates with patient size as mea- pending on maternal size. authors; literature research, E.A., M.M.G., D.D.C., C.H.M.,
sured by the perimeter around 䡲 Fetal dose estimates for a single- M.F.M.; experimental studies, E.A., C.V.W., M.M.G., N.Y.,
the mother. pass abdominal and pelvic acqui- J.J.D., C.H.C., D.D.C., D.M.S., A.N.P., M.F.M.; statistical
䡲 There was no correlation between sition are below the consensus analysis, E.A., M.M.G., J.W.S., M.F.M.; and manuscript
gestational age and fetal dose levels for negligible risk (50 –150 editing, E.A., C.V.W., M.M.G., J.J.D., C.H.C., D.D.C.,
D.M.S., A.N.P., C.H.M., M.F.M.
(P ⫽ .543). mGy) and well below the action-
䡲 The average fetal doses were esti- able level of 150 mGy. Funding:
mated to be 10.8 mGy per 100 䡲 Modulation of tube current with This research was funded by the National Institute of
mAs for a specific scanner model, respect to patient size should be Biomedical Imaging and Bioengineering (grants R01
EB004898, T32 EB002101).
with a range from 7.3 to 14.3 used to decrease fetal dose for
mGy/100 mAs. smaller maternal sizes. Authors stated no financial relationship to disclose.

Radiology: Volume 249: Number 1—October 2008 ▪ radiology.rsnajnls.org 221

MEDICAL PHYSICS: Monte Carlo Simulation in Pregnant Patient Models Angel et al

the basis of image data from a cohort of mm. The patients in the cohort were patient data by one of two board-certi-
31 pregnant patients who previously un- originally scanned with one of the fied radiologists (C.V.W., N.Y.). The
derwent clinically indicated abdominal following scanners: HighSpeed CT/I, radiologist contoured the uterus, gesta-
and pelvic CT examination. The original LightSpeed QX/i, LightSpeed Ultra, tional sac (when visible), and fetus
CT examinations were performed unre- LightSpeed PRO, LightSpeed 16, (when visible), which resulted in a set of
lated to this study and were often in or LightSpeed VCT (GE Healthcare, Wau- voxels that were labeled as belonging to
emergency and/or trauma patients (such kesha, Wis). Oral contrast agent had these three tissue groups. Images were
as those who were injured in a motor been used in 13 of the 24 CT examina- subsampled from a 512 ⫻ 512 matrix to
vehicle accident). Institutional review tions. The cranial-caudal length of each a 128 ⫻ 128 matrix to reduce simulation
board approval was obtained for this voxelized patient model was dependent time.
Health Insurance Portability and Ac- on the length of CT image data avail- Voxels within the fetus were mod-
countability Act– compliant study for able. The average length of image data eled as soft tissue or soft bone, de-
the use of the anonymized CT images, was 46.2 cm, and all of the image series pending on the CT number (in
as well as for additional anonymized pa- included, at a minimum, the patient Hounsfield units). The voxels in the
tient data (ie, patient weight and gesta- anatomy from the lower thorax to the gestational sac were modeled as wa-
tional age). The 24 patient models were pubic symphysis. ter, and the voxels in the uterus were
selected to represent a range of gesta- Two anatomic measures were ob- modeled as soft tissue. Each voxel of
tional ages; seven cases were excluded tained from each patient’s images by one the mother outside these fetal regions
because of a lack of sufficient volumetric of the authors (E.A.), who had been was modeled as one of six tissue types
image data (images did not fully cover trained in reviewing CT image data. (lung, fat, water, muscle, bone, or air)
the uterus) or because gestational ages These measures were used to assess the (13). The process of creating voxel-
were similar to those of other cases al- size of the mother and the location of the ized models from the original patient
ready in the cohort. The set of 24 pa- fetus within the mother. The patient size images is illustrated in Figure 3.
tients included gestational ages ranging metric was the outer (skin) perimeter of
from less than 5 weeks to 36 weeks, the patient on the image containing the Multidetector CT Source Model
with an average gestational age of 20 three-dimensional geometric centroid of A previously validated multidetector CT
weeks. The cohort included six patients the fetus. The perimeter was determined source model was created by using
in the first trimester of pregnancy, 10 by using a semiautomated threshold- Monte Carlo simulation– based tech-
patients in the second trimester, and based contouring technique. Fetal depth niques (14–18). The CT scanner mod-
eight patients in the third trimester. Ex- was measured manually as the distance eled was a LightSpeed 16 (GE Medical
amples of these CT images are shown in from the anterior skin surface to the most Systems). The scanner’s x-ray spec-
Figure 1. The 24 models created as part anterior part of the fetus. The patient pe- trum, filtration, and geometry were ob-
of this study are publicly available at rimeter and fetal depth measures are il- tained from the manufacturer and con-
http://medqia.org/⬃mcnitt/FetalModels. lustrated in Figure 2. firmed through measurements. The x-ray
The original image data were acquired spectrum was initially described as a func-
between January 2002 and February 2006 Patient Models tion of the number of photons at 1-keV
with 120-kVp acquisition protocols, with To estimate fetal dose, three regions of energy intervals, which was then con-
image thickness varying from 1.25 to 10 interest were identified for each set of verted to a cumulative distribution func-

Figure 1

Figure 1: CT images in three patients in the cohort at gestational age of 7 weeks (left), 24 weeks (middle), and 36 weeks (right) that demonstrate early term, midterm,
and late-term pregnancy, respectively.

222 radiology.rsnajnls.org ▪ Radiology: Volume 249: Number 1—October 2008

MEDICAL PHYSICS: Monte Carlo Simulation in Pregnant Patient Models Angel et al

Figure 2
tion for implementation within the which resulted in a first half-value layer
Monte Carlo source model (14–18). at isocenter of approximately 7.8 mm
The multidetector CT model included aluminum. The simulated scan had a
the bowtie filter and beam collima- helical source path with a pitch of 1.0
tion. The multidetector CT model was and had 16 detector rows of 1.25 mm
designed to simulate acquisition param- (resulting in 16 ⫻ 1.25-mm total nomi-
eters selected by the user, including nal collimation). Although the actual
beam energy (kilovoltage peak), source scanner modeled does not allow these
path (ie, helical or axial), pitch, total particular acquisition parameters (pitch
nominal beam collimation, and scan of 1.0 is not available), the simulation
length. parameters represent a typical abdomi-
nal and pelvic examination, and the re-
Monte Carlo Simulation of Abdominal and sults can be scaled to any other pitch
Pelvic CT Examination value. Simulated examinations were per-
Figure 2: CT image with semiautomated con- The simulated CT examination covered the formed in all 24 voxelized phantoms. No
tour (dashed line) around the perimeter of the abdominal and pelvic region. The total tube current–time product value was
mother on central image in a 35-week-old fetus, as length of the simulated examination (in- specified; all doses were tallied and re-
well as the assessment of fetal depth (distance cluding image data and z-axis overscan) ported on a normalized basis of milli-
from skin to fetus [double arrow]). was equivalent to the length of the orig- grays per 100 mAs.
inal patient image data. The simulated The simulation involves modeling
abdominal and pelvic examination had a the x-ray photon transport through the
beam energy and filtration combination voxelized patients and tallying dose in
of 120 kVp and body bow tie filter, the voxels that were labeled as uterus,

Figure 3

Figure 3: CT images and voxelized models created from patient images of an early pregnancy (7 weeks) and a later pregnancy (35 weeks). Pink represents the uterus,
yellow represents the gestational sac, and red represents the fetus (when visible). In the voxelized model, fetal bone is white, and tissues outside the radiologist’s con-
tours range from blue to green.

Radiology: Volume 249: Number 1—October 2008 ▪ radiology.rsnajnls.org 223

MEDICAL PHYSICS: Monte Carlo Simulation in Pregnant Patient Models Angel et al

Table 1
gestational sac, and fetus. All simula-
tions used the Monte Carlo N-Particle Patient Model Measurements and Radiation Dose per 100 mAs to Fetus according to
eXtended code (version 2.5.0; Radia- Gestational Age of 24 Voxelized Models
tion Safety Information Computational
Normalized Fetal Dose
Center, Oak Ridge, Tenn) created at Gestational Age (wk) Maternal Perimeter (cm) Fetal Depth (cm) (mGy/100 mAs)
the Los Alamos National Laboratory
(19,20). Each Monte Carlo simulation ⬍5.0 123 10.6 7.3*
was performed with 1–10 million simu- 5.0 89 4.2 11.8†
lated x-ray photons, which were enough 5.0 88 7.6 10.3†
to ensure less than 1% statistical error 6.6 102 10.9 8.8†
within the tally region of interest (ap- 7.1 90 5.9 12.6†
12.1 88 4.6 14.2
proximately 2 hours run time on an av-
14.3 105 6.5 10.9
erage desktop computer).
14.9 93 7.1 11.2
Dose Calculations 17.0 94 7.7 11.3
17.1 87 6.7 12.2
The average radiation dose to the fetus
18.5 87 5.6 14.3
was tallied when the fetus was distin- 20.3 112 8.0 8.5
guishable from the gestational sac and 22.0 108 4.7 11.1
uterus. If the fetus was not visible, the 23.7 118 6.3 8.1
average dose was tallied in the gesta- 24.0 95 5.6 11.5
tional sac. In one voxelized model, with 24.4 94 6.6 11.5
a gestational age less than 5 weeks, the 25.0 92 2.5 12.3
gestational sac was not visible; in this 27.0 89 9.0 11.2
case, the dose was tallied in the uterus. 27.4 104 3.6 11.9
The average dose was computed by av- 27.4 122 6.0 8.6
eraging the values of dose absorbed in 28.3 119 5.5 9.5
the voxels within the tally region. Ab- 29.4 105 3.5 9.7
sorbed dose within a voxel was com- 35.0 100 5.1 10.4
puted from the collision kerma, calcu- 35.9 119 3.4 8.9
lated from the Monte Carlo N-Particle Note.—The means for gestational age, maternal perimeter, fetal depth, and normalized fetal dose were 19.7 weeks, 101 cm,
eXtended track-length estimate of en- 6.1 cm, and 10.8 mGy/100 mAs, respectively.
ergy fluence, and multiplied by the ma- * Uterus dose was calculated because the fetus and gestational sac were not visible.
terial-specific and energy-dependent †
Gestational sac dose was calculated because the fetus was not visible.
mass-energy absorption coefficient. The
mass-energy absorption coefficients were
taken from the tables of Hubbell and
Seltzer (21). All reported dose results
were normalized per 100 mAs. Table 2

Radiation Dose to Fetus as Estimated with Monte Carlo Technique Compared with
Statistical Analysis That of Three Existing Fetal Dose Estimation Techniques
The fetal dose estimates from the co-
Normalized Fetal Dose
hort of voxelized patient models were
(mGy/100 mAs)
tested for correlation with each of three
Dose Estimation Method No. of Patient Models Average Range
independent measures described above:
gestational age, perimeter of the Monte Carlo 24 10.8 7.3–14.3
mother, and fetal depth. Stepwise mul- Felmlee et al (1) (recommended ⫾20% size range) 1 11.3 9.0–13.6
tiple linear regression analysis was per- ImPACT 1 12.0 NA
formed to assess the partial correlation CT-Expo 1 11.7 NA
of each variable, and P values were cal- Note.—NA ⫽ not applicable.
culated to determine statistical signifi-
cance. A P value less than .05 indicated
a significant difference. this study. However, there are several estimation of fetal dose was that of Felm-
existing fetal dose estimation methods lee et al (1). For these calculations, the
Existing Methods against which the results of this study fetus was assumed to be axially centered,
There is no reference standard for fetal can be compared. and contiguous axial scans were used in-
dose that can be used as comparison for The first comparison method for the stead of helical scans, with a pitch of 1.0.

224 radiology.rsnajnls.org ▪ Radiology: Volume 249: Number 1—October 2008

MEDICAL PHYSICS: Monte Carlo Simulation in Pregnant Patient Models Angel et al

Although this technique does not address These methods were used to estimate regression analysis (R2 ⫽ 0.017) indi-
gestational age and produces a single fetal dose to the uterus and represent fetal cated no significant correlation (r ⫽
dose result, Felmlee et al state that a dose dose for gestational ages approximately ⫺0.130; P ⫽ .543) between gestational
estimate may vary by ⫾20% due to varia- less than or equal to 8 weeks in a stan- age and fetal dose.
tions in patient size. The estimated fetal dard-size patient. The perimeter of the Fetal dose as a function of the pe-
dose and a ⫾20% range were compared hermaphrodite MIRD phantom used with rimeter of the mother is shown in
with the Monte Carlo results from this the ImPACT software was approximately Figure 5; the relationship between the
study. The abdomen of the anthropo- 99 cm at the pelvis (6). The Eva geomet- perimeter (P) (in centimeters) of the
morphic phantom used for the Felmlee ric phantom used with the CT-Expo soft- mother and fetal dose (DOf) (in milli-
technique had a perimeter of approxi- ware has an anteroposterior size of 18.8 grays per 100 mAs) can be repre-
mately 88 cm. cm (7) and an estimated perimeter of 93 sented as:
The second and third fetal dose esti- cm. The ImPACT and CT-Expo software
mation methods used for comparison each produced a single fetal dose esti- DOf ⫽ ⫺0.122(P) ⫹ 23.11, (1)
were the ImPACT dose calculator (4) and mate, which was compared with the re-
CT-Expo software (7,8), which are both sults from this study. which has an R2 value of 0.681, indicat-
based on Monte Carlo simulations (5,10). For each dose comparison method, ing a significant linear correlation (r ⫽
the CT acquisition parameters (tube ⫺0.825; P ⬍ .001) between maternal
voltage, collimation, etc) were matched size and fetal dose. The standard error
Figure 4
as closely as possible to the protocol of the estimate for the perimeter model
simulated in this study. The scan length was 1.053.
for the abdominal and pelvic examina- A two-variable model including pa-
tion was selected anatomically to span tient perimeter (in centimeters) and fe-
from the lower thorax to inferior of the tal depth (DEf) (in centimeters) pro-
pubic symphysis. This corresponded to vides a better correlation with fetal dose
a 50-cm scan length for the Felmlee (milligrays per 100 mAs), with an R2
technique, a 50-cm scan length for the value of 0.799 (r ⫽ ⫺0.894; P ⫽ .002).
ImPACT technique, and a 48-cm scan The two-variable model allows fetal
length for the CT-Expo technique. dose to be estimated by:

DOf ⫽ ⫺0.119(P) ⫺ 0.029(DEf)

Results
Figure 4: Graph of radiation dose to fetus ver- ⫹ 24.56. (2)
sus gestational age. Gestational age does not sig-
Dose Calculation The standard error of the estimate
nificantly correlate with normalized fetal dose
(R 2 ⫽ 0.017; P ⫽ .543). Table 1 shows the gestational age, pe- for the two-variable model was 0.855.
rimeter of the mother, fetal depth, and
the estimated fetal dose for each of the
24 patient models. The average fetal
Discussion
dose was 10.8 mGy/100 mAs ⫾ 1.8 Radiation dose to the fetus from mul-
Figure 5
(standard deviation). The minimum, tidetector CT imaging in pregnant pa-
maximum, and median fetal doses were tients was estimated for a range of
7.3, 14.3, and 11.2 mGy/100 mAs, re- actual pregnant patient anatomies,
spectively. yielding a mean value of 10.8 mGy/100
mAs with a range of 7.3–14.3 mGy/
Comparison with Existing Methods 100 mAs.
The results of the existing methods are Patient size (perimeter) had a signif-
shown in Table 2; the Felmlee et al icant correlation with fetal dose (P ⬍
method produced estimates of 11.3 .001). This size metric was chosen be-
mGy/100 mAs (the ⫾20% range being cause it can be easily measured in the
equal to 9.0 –13.6), while the ImPACT clinic by using a tape measure around
and CT-Expo methods yielded estimates the estimated center of the fetus. The
of 12.0 and 11.7 mGy/100 mAs, respec- fetal depth also correlated well with fe-
Figure 5: Graph of radiation dose to fetus ver- tively. tal dose, resulting in a two-variable
sus perimeter of mother. Perimeter of mother cor- model for the estimation of fetal dose.
relates linearly with normalized fetal dose (R2 ⫽ Regression Analysis Fetal depth may be measured on an ul-
0.681; P ⬍ .001; standard error of the estimate ⫽
Fetal dose as a function of gestational trasonographic image of the patient or,
1.053).
age alone is shown in Figure 4. Linear if available, directly from a CT image.

Radiology: Volume 249: Number 1—October 2008 ▪ radiology.rsnajnls.org 225

MEDICAL PHYSICS: Monte Carlo Simulation in Pregnant Patient Models Angel et al

There was no correlation between ges- should be scaled proportional to the in- only one multidetector CT scanner model
tational age and fetal dose. There was verse of pitch (22). For example, to es- was used in this study. We expect that
insufficient information available on the timate the dose for an examination with there may be substantial differences in
height or weight of patients to investi- a pitch of 1.375 (with all other technical radiation dose per unit of tube current–
gate correlations of these patient size parameters constant), the results from time product with the various makes
metrics with fetal dose. this study can be divided by 1.375. and models of CT scanners due to dif-
When we compared existing dose To illustrate a dose calculation using ferences in x-ray beam filtration
estimation methods with the results of all of the parameters that were found to schemes. The differences between
this study, we found that the Felmlee be important in this study, consider an these schemes may not be easily taken
et al technique for the estimation of ra- abdominal and pelvic helical CT exami- into account; therefore, we are cur-
diation dose produced a good estimate nation in a pregnant patient with a pe- rently extending our Monte Carlo
of fetal dose for an average-size patient. rimeter of 101 cm and a fetal depth of 7 source models to those from other man-
The ImPACT and CT-Expo fetal dose cm with the following acquisition proto- ufacturers. This study also simulated
estimates were slightly conservative be- col: GE LightSpeed 16, 120 kVp, pitch only constant tube current scans and
cause fetal dose was slightly overesti- of 1.375, 300 mAs (220 effective mAs), therefore did not assess any dose sav-
mated relative to that estimated with and 16 ⫻ 1.25-mm total nominal colli- ings that may be offered by using tube
the methods and patient cohort used in mation. By substituting the mother’s pe- current modulation methods. This is
this study. The estimates with the rimeter and the fetal location in Equa- also the subject of ongoing evaluations.
ImPACT and CT-Expo methods were tion (2) and multiplying the result by a A method for the estimation of fetal
still within the range of the doses esti- factor of 3 (300 mAs divided by 100 dose from models of actual patient anat-
mated for the patients modeled in this mAs) and dividing by the 1.375 pitch, omy that represented a range of gesta-
study. one obtains a fetal dose of approxi- tional age and patient size was devel-
Of note, however, is that the results mately 27 mGy. (Note that the gesta- oped. The mean fetal dose estimated
demonstrate a range of fetal doses that tional age is not a factor in this esti- across the patient cohort was 10.8
varied by nearly a factor of two (7.3– mate.) mGy/100 mAs, which is consistent with
14.3 mGy/100 mAs) among the 24 pa- These results indicate that, while estimates from existing methods. How-
tients modeled. This difference in fetal there may be wide variation among indi- ever, the results from this study show a
dose exceeded the range suggested by vidual patients, fetal doses from a single larger spread of values than previously
Felmlee et al. abdominal and pelvic CT examination recognized, thus demonstrating the im-
The results from this study are re- are still below the consensus levels or portance of accounting for patient size
ported as normalized dose per 100 mAs negligible risk (50 –150 mGy) and well in fetal dose estimates. Fetal dose cor-
to allow for dose estimations of exami- below the “actionable” level of 150 mGy related with maternal perimeter, and
nations with any tube current–time (23). This should not be interpreted as this correlation improved when both
product. For example, for an examina- CT imaging in pregnant patients should maternal size and fetal depth were ac-
tion with 300 mAs, the normalized re- be performed without adequate clinical counted for.
sults from this study should be tripled. justification. It must also be considered Acknowledgment: This work was made possi-
The results can also be used to estimate that the dose estimates from this study ble by assistance from Mohammad Ghatali, MD.
fetal dose for CT protocols utilizing size- represent a single CT acquisition. Many
dependent tube current–time product clinical examination protocols require
References
settings. For example, the results for two or three acquisitions (ie, pre- and
1. Felmlee JP, Gray JE, Leetzow ML, Price JC.
pregnant patients with larger perime- postcontrast imaging), thus doubling or
Estimated fetal radiation dose from multi-
ters can be scaled according to the tripling the fetal dose. slice CT studies. AJR Am J Roentgenol 1990;
tube current–time product settings for There were several limitations to 154(1):185–190.
heavier or larger patients, and the re- this study. While these methods did in-
2. Shope TB, Gagne RM, Johnson GC. A
sults for patients with smaller perime- clude an extensive set of voxelized pa-
method for describing the doses delivered by
ters can be scaled according to the tube tient models of pregnant patients, they transmission x-ray computed tomography.
current–time product settings for smaller may not be representative of the gen- Med Phys 1981;8(4):488 – 495.
patients. eral population of all pregnant patients.
3. Hurwitz LM, Yoshizumi T, Reiman RE, et al.
Although the Monte Carlo simula- While the cohort of patients used in this
Radiation dose to the fetus from body MDCT
tions in this study are based on a single study spanned a range of gestational during early gestation. AJR Am J Roentgenol
set of acquisition parameters, the re- ages, the range of patient sizes encoun- 2006;186(3):871– 876.
sults can be used to estimate dose for tered in clinical practice may be wider
4. ImPACT CTDosimetry. Imaging perfor-
various protocols that differ in pitch, than those of the group presented here,
mance assessment of CT scanners: a medi-
tube current (in milliamperes), or gan- and because perimeter correlated sig- cal devices agency evaluation group. CT
try rotation time (in seconds). To ac- nificantly with fetal dose, this could be scanner matching data, tables of CTDI val-
count for differences in pitch, dose important. Another limitation was that ues in air, CTDIw, and phantom factor val-

226 radiology.rsnajnls.org ▪ Radiology: Volume 249: Number 1—October 2008

MEDICAL PHYSICS: Monte Carlo Simulation in Pregnant Patient Models Angel et al

ues. ImPACT Internet home page. http: tion of dose from external photon exposures amination. Phys Med Biol 2006;51(20):
//www.ImPACTscan.org. Accessed August using reference human phantoms and Monte 5151–5166.
2000. Carlo methods. VI. Organ doses from com-
17. Demarco JJ, Cagnon CH, Cody DD, et al.
puted tomographic examinations. GSF re-
5. Jones DG, Shrimpton PC. Survey of the Estimating radiation doses from multidetec-
port 30/91. Oberschleissheim, Germany:
practice in the UK. III. Normalized organ tor CT using Monte Carlo simulations: ef-
GSF-Forschungszentrum, 1991.
doses calculated using Monte Carlo tech- fects of different size voxelized patient mod-
niques. NRPB R-250. Chilton, England: Na- 11. Shi C, Xu XG. Development of a 30-week- els on magnitudes of organ and effective
tional Radiological Protection Board, 1991. pregnant female tomographic model from dose. Phys Med Biol 2007;52(9):2583–2597.

6. Snyder WS, Fisher HL, Ford MR, et al. Esti- computed tomography (CT) images for 18. Lee C, Lee C, Staton RJ, et al. Organ and
mates of absorbed fractions for monoener- Monte Carlo organ dose calculations. Med effective doses in pediatric patients undergo-
getic photon sources uniformly distributed in Phys 2004;31(9):2491–2497. ing helical multislice computed tomography
various organs of a heterogeneous phantom. examination. Med Phys 2007;34(5):1858 –
12. Osei EK, Faulkner K. Fetal position and size
J Nucl Med 1969;(supp 3):7–52. 1873.
data for dose estimation. Br J Radiol 1999;
7. Stamm G, Nagel HD. CT-expo: a novel pro- 72(856):363–370. 19. Waters L, ed. 2002 MCNPX user’s manual,
gram for dose evaluation in CT [in German]. version 2.4.0. Los Alamos National Labora-
13. DeMarco JJ, Solberg TD, Smathers JB. A tory report LA-CP-02– 408. 2002.
Rofo 2002;174(12):1570 –1576.
CT-based Monte Carlo simulation tool for
8. Nagel HD. Radiation exposure in computed dosimetry planning and analysis. Med Phys 20. Waters L, ed. 2003 MCNPX user’s manual,
tomography, fundamentals, influencing pa- 1998;25(1):1–11. version 2.5.C. Los Alamos National Labora-
rameters, dose assessment, optimisation, tory report LA-UR-03–2202. 2003.
scanner data, terminology. COCOR Euro- 14. Jarry G, DeMarco JJ, Beifuss U, Cagnon CH,
21. Hubbell JH, Seltzer SM. Tables of x-ray mass
pean Coordination Committee of the Radio- McNitt-Gray MF. A Monte Carlo-based
absorption coefficients and mass energy-ab-
logical and Electromedical Industries, 2nd ed method to estimate radiation dose from spi-
sorption coefficients (version 1.03). http:
(in English, revised and translated by Nagel ral CT: from phantom testing to patient-spe-
//physics.nist.gov/. Gaithersburg, Md: Na-
HD and Shrimpton PC). Hamburg, Germany: cific models. Phys Med Biol 2003;48(16):
tional Institute of Standards and Technology.
Paul Hartung Druck, 2000. 2645–2663.
1995.
9. Kramer R, Zankl M, Williams G, Drexler G. 15. DeMarco JJ, Cagnon CH, Cody DD, et al. A 22. McNitt-Gray MF. AAPM/RSNA physics tu-
The calculation of dose from external photon Monte Carlo based method to estimate radi- torial for residents: topics in CT—radia-
exposures using reference human phantoms ation dose from multidetector CT (MDCT): tion dose in CT. RadioGraphics 2002;
and Monte Carlo methods. I. The male cylindrical and anthropomorphic phantoms. 22(6):1541–1553.
(Adam) and female (Eva) adult mathemati- Phys Med Biol 2005;50(17):3989 – 4004.
23. McCollough CH, Schueler BA, Atwell TD,
cal phantoms. Neuherberg, Germany: GSF-
16. Staton RJ, Lee C, Lee C, et al. Organ and et al. Radiation exposure and pregnancy:
Bericht, 1982;S-885.
effective doses in newborn patients during when should we be concerned? Radio-
10. Zankl M, Panzer W, Drexler G. The calcula- helical multislice computed tomography ex- Graphics 2007;27(4):909 –917.

Radiology: Volume 249: Number 1—October 2008 ▪ radiology.rsnajnls.org 227

 Radiology 2008
This is your reprint order form or pro forma invoice
(Please keep a copy of this document for your records.)

Reprint order forms and purchase orders or prepayments must be received 72 hours after receipt of form either
by mail or by fax at 410-820-9765. It is the policy of Cadmus Reprints to issue one invoice per order.
Please print clearly.

Author Name _______________________________________________________________________________________________

Title of Article _______________________________________________________________________________________________
Issue of Journal_______________________________ Reprint # _____________ Publication Date ________________
Number of Pages_______________________________ KB # _____________ Symbol Radiology
Color in Article? Yes / No (Please Circle)
Please include the journal name and reprint number or manuscript number on your purchase order or other correspondence.
Order and Shipping Information

Reprint Costs (Please see page 2 of 2 for reprint costs/fees.) Shipping Address (cannot ship to a P.O. Box) Please Print Clearly
Name ___________________________________________
________ Number of reprints ordered $_________ Institution _________________________________________
________ Number of color reprints ordered $_________ Street ___________________________________________
________ Number of covers ordered $_________ City ____________________ State _____ Zip ___________
Country ___________________________________________
Subtotal $_________ Quantity___________________ Fax ___________________
Taxes $_________ Phone: Day _________________ Evening _______________
(Add appropriate sales tax for Virginia, Maryland, Pennsylvania, and the E-mail Address _____________________________________
District of Columbia or Canadian GST to the reprints if your order is to
be shipped to these locations.) Additional Shipping Address* (cannot ship to a P.O. Box)
Name ___________________________________________
First address included, add $32 for Institution _________________________________________
each additional shipping address $_________ Street ___________________________________________
City ________________ State ______ Zip ___________
Country _________________________________________
Quantity __________________ Fax __________________
TOTAL $_________ Phone: Day ________________ Evening ______________
E-mail Address ____________________________________
* Add $32 for each additional shipping address

Payment and Credit Card Details Invoice or Credit Card Information

Enclosed: Personal Check ___________ Invoice Address Please Print Clearly
Please complete Invoice address as it appears on credit card statement
Credit Card Payment Details _________ Name ____________________________________________
Checks must be paid in U.S. dollars and drawn on a U.S. Bank.
Institution ________________________________________
Credit Card: __ VISA __ Am. Exp. __ MasterCard Department _______________________________________
Card Number __________________________________ Street ____________________________________________
Expiration Date_________________________________ City ________________________ State _____ Zip _______
Signature: _____________________________________ Country ___________________________________________
Phone _____________________ Fax _________________
Please send your order form and prepayment made payable to:
E-mail Address _____________________________________
Cadmus Reprints
P.O. Box 751903 Cadmus will process credit cards and Cadmus Journal
Charlotte, NC 28275-1903 Services will appear on the credit card statement.
Note: Do not send express packages to this location, PO Box.
If you don’t mail your order form, you may fax it to 410-820-9765 with
FEIN #:541274108 your credit card information.
Signature __________________________________________ Date _______________________________________
Signature is required. By signing this form, the author agrees to accept the responsibility for the payment of reprints and/or all charges
described in this document.
RB-9/26/07
Page 1 of 2
 Radiology 2008
Black and White Reprint Prices Color Reprint Prices
Domestic (USA only) Domestic (USA only)
# of # of
50 100 200 300 400 500 50 100 200 300 400 500
Pages Pages
1-4 $221 $233 $268 $285 $303 $323 1-4 $223 $239 $352 $473 $597 $719
5-8 $355 $382 $432 $466 $510 $544 5-8 $349 $401 $601 $849 $1,099 $1,349
9-12 $466 $513 $595 $652 $714 $775 9-12 $486 $517 $852 $1,232 $1,609 $1,992
13-16 $576 $640 $749 $830 $912 $995 13-16 $615 $651 $1,105 $1,609 $2,117 $2,624
17-20 $694 $775 $906 $1,017 $1,117 $1,220 17-20 $759 $787 $1,357 $1,997 $2,626 $3,260
21-24 $809 $906 $1,071 $1,200 $1,321 $1,471 21-24 $897 $924 $1,611 $2,376 $3,135 $3,905
25-28 $928 $1,041 $1,242 $1,390 $1,544 $1,688 25-28 $1,033 $1,071 $1,873 $2,757 $3,650 $4,536
29-32 $1,042 $1,178 $1,403 $1,568 $1,751 $1,924 29-32 $1,175 $1,208 $2,122 $3,138 $4,162 $5,180
Covers $97 $118 $215 $323 $442 $555 Covers $97 $118 $215 $323 $442 $555

International (includes Canada and Mexico) International (includes Canada and Mexico))
# of # of
50 100 200 300 400 500 50 100 200 300 400 500
Pages Pages
1-4 $272 $283 $340 $397 $446 $506 1-4 $278 $290 $424 $586 $741 $904
5-8 $428 $455 $576 $675 $784 $884 5-8 $429 $472 $746 $1,058 $1,374 $1,690
9-12 $580 $626 $805 $964 $1,115 $1,278 9-12 $604 $629 $1,061 $1,545 $2,011 $2,494
13-16 $724 $786 $1,023 $1,232 $1,445 $1,652 13-16 $766 $797 $1,378 $2,013 $2,647 $3,280
17-20 $878 $958 $1,246 $1,520 $1,774 $2,030 17-20 $945 $972 $1,698 $2,499 $3,282 $4,069
21-24 $1,022 $1,119 $1,474 $1,795 $2,108 $2,426 21-24 $1,110 $1,139 $2,015 $2,970 $3,921 $4,873
25-28 $1,176 $1,291 $1,700 $2,070 $2,450 $2,813 25-28 $1,290 $1,321 $2,333 $3,437 $4,556 $5,661
29-32 $1,316 $1,452 $1,936 $2,355 $2,784 $3,209 29-32 $1,455 $1,482 $2,652 $3,924 $5,193 $6,462
Covers $156 $176 $335 $525 $716 $905 Covers $156 $176 $335 $525 $716 $905

Minimum order is 50 copies. For orders larger than 500 copies, Tax Due
please consult Cadmus Reprints at 800-407-9190. Residents of Virginia, Maryland, Pennsylvania, and the District
of Columbia are required to add the appropriate sales tax to each
Reprint Cover reprint order. For orders shipped to Canada, please add 7%
Cover prices are listed above. The cover will include the Canadian GST unless exemption is claimed.
publication title, article title, and author name in black.
Ordering
Reprint order forms and purchase order or prepayment is
Shipping required to process your order. Please reference journal name
Shipping costs are included in the reprint prices. Domestic and reprint number or manuscript number on any
orders are shipped via UPS Ground service. Foreign orders are correspondence. You may use the reverse side of this form as a
shipped via a proof of delivery air service. proforma invoice. Please return your order form and
prepayment to:
Multiple Shipments
Cadmus Reprints
Orders can be shipped to more than one location. Please be
P.O. Box 751903
aware that it will cost $32 for each additional location.
Charlotte, NC 28275-1903
Delivery Note: Do not send express packages to this location, PO Box.
Your order will be shipped within 2 weeks of the journal print FEIN #:541274108
date. Allow extra time for delivery.
Please direct all inquiries to: Reprint Order Forms
and purchase order
Rose A. Baynard or prepayments must
800-407-9190 (toll free number) be received 72 hours
410-819-3966 (direct number) after receipt of form.
410-820-9765 (FAX number)
[email protected] (e-mail)

Page 2 of 2