| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, SUNY Upstate Medical University, 750 E Adams St.,
Syracuse, NY 13210-2306.
2 Present address: Department of Diagnostic Imaging, Rhode Island Hospital,
Providence, RI 02903.
Received January 19, 2006;
accepted after revision June 7, 2006.
Address correspondence to W. Huda
(hudaw{at}upstate.edu).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Heads, chests, and abdomens of patients ranging from neonates to oversized adults (120 kg) were modeled as uniform cylinders of water. Monte Carlo dosimetry data were used to obtain average doses in the directly irradiated region. Dosimetry data were used to compute the total energy imparted, which was converted into the corresponding effective dose using patient-size-dependent effective-dose-per-unit-energy-imparted coefficients. Representative patient doses were obtained for scanning protocols that take into account the size of the patient being scanned by typical MDCT scanners.
RESULTS. Relative to CT scanners from the early 1990s, present-day
MDCT scanners result in doses that are 
1.5 and 1.7 higher per unit
mAs in head and body phantoms, respectively. Organ absorbed doses in head CT
scans increase from 30 mGy in newborns to 40 mGy in adults. Patients
weighing less than 20 kg receive body organ absorbed doses of 7 mGy,
which is a factor of 2 less than for normal-sized (70-kg) adults. Adult head
CT effective doses are 0.9 mSv, four times less than those for the
neonate. Effective doses for neonates undergoing body CT are 2.5 mSv,
whereas those for normal-sized adults are 3.5 mSv.
CONCLUSION. Representative organ absorbed doses in CT are substantially lower than threshold doses for the induction of deterministic effects, and effective doses are comparable to annual doses from natural background radiation.
Keywords: MDCT • pediatric CT • physics of radiology • radiation dose
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The dose characteristics of new MDCT scanners merit investigation to give practitioners a better understanding of how radiation doses from these newer systems compare with the single-detector systems prevalent in the 1990s. A major effort has recently been undertaken to optimize CT [8, 9] using imaging protocols that explicitly take into account the characteristics of the patient being examined [10-12]. It is therefore of interest to investigate how the introduction of MDCT scanners and patient-size-dependent imaging protocols have affected patient doses.
Patients undergoing CT examinations can range from neonates to oversized adults. Radiation doses in CT, however, are generally measured in cylindric acrylic phantoms designed to simulate a head (16-cm diameter) or body (32 cm) [13]. It is difficult to obtain reliable quantitative values of patient doses from any measurements performed in these standard dose phantoms because patients have sizes and body compositions that can differ markedly from the phantoms (e.g., pediatric patients). In this article, we modeled patients as uniform cylinders of water to estimate representative CT doses for current generation MDCT scanners. We investigated patients ranging from the neonate to the oversized adult who are scanned using protocols with radiographic techniques that explicitly account for patient size.
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Data on head characteristics as a function of patient age have been published for patients ranging from the neonate to the adult [17]. The third column in Table 1 summarizes the radii of water cylinders that can be used to simulate heads of patients ranging from the neonate to the adult. Patient body dimensions and the corresponding average Hounsfield values have been published by Ogden et al. [18]. Column 4 of Table 1 summarizes the resultant water equivalent radii for the chest region used in this study, and column 5 summarizes the corresponding data for the abdomen.
|
Dosimetry
Organ absorbed doses—When a water cylinder (radius
r) located at a CT scanner isocenter undergoes a single rotation of
the X-ray tube (i.e., 360°) at a fixed technique value (kV/mAs),
approximately half the energy will be deposited in the directly irradiated
region and the remaining half in scatter tails on either side of the directly
irradiated region [19]. The
mean section dose, Dm, is defined as the total energy
deposited in the water cylinder divided by the mass of the directly irradiated
volume, or 
r2 T 
, where T is the
section thickness and is the water density
[20]. The mean section dose
approximates the average (i.e., representative) dose in the directly
irradiated region of the water cylinder for contiguous scanning in the axial
scanning mode, or using a pitch ratio of 1 in the helical scanning mode.
Figure 1 shows how the mean
section dose, Dm, varies with water cylinder radius
obtained using Monte Carlo modeling techniques for a HiSpeed Advantage CT
scanner (GE Healthcare) operated at 120 kV
[20]. Increasing the cylinder
radius from 20 to 200 mm reduces the mean section dose from 0.185 to 0.030
mGy/mAs, or by a factor of about 6. Figure
2 shows the relative mean section dose, RkV,
as a function of X-ray tube voltage (constant mAs), and normalized to unity at
120 kV for each phantom size. Increasing the X-ray tube voltage from 80 to 140
kV at a fixed mAs increases Dm by a factor of about 5.5.
For the HiSpeed Advantage scanner at given kV and mAs values, representative
organ doses, Dorgan, may be obtained using the equation
 | (1) |
where P is the pitch ratio and RkV is taken from the data shown in Figure 2. The HiSpeed Advantage was introduced into clinical practice in the early 1990s and is not representative of current MDCT scanners. Table 2 summarizes data for head- and body-weighted CT dose index (CTDIw) data for recent CT systems from five major vendors [13]. The data presented in Table 2 indicate that head and body doses on presentday scanners (mGy/mAs), when operated at 120 kV, would be 1.45 and 1.71 times higher, respectively, than those shown in Figure 1.
|
|
), with a mean section dose Dm per 360°
rotation of the X-ray tube, is
 | (2) |
where T is the nominal X-ray beam width and N is the total number of rotations of the X-ray tube around the phantom. For a CT scanner operated in the helical mode, N may be estimated by dividing the full scan length by the product of the X-ray beam width and the pitch ratio P.
For a given patient size and scanned body region, the effective dose is
directly proportional to the energy imparted to the patient. Doubling the
energy imparted to a patient undergoing head CT, for example, is generally
expected to double the corresponding effective dose. Published Monte Carlo
calculations of radiation doses at CT examinations provide the complete
pattern of energy deposition in the patient and enable the mean dose to each
organ and tissue to be determined. These organ doses enable the computation of
the effective dose (E) and the energy imparted ()
[20], where the ratio of these
two parameters, E / , permits energy imparted to be converted
into a corresponding effective dose using
 | (3) |
where (E / )R,M is the effective dose
per unit energy imparted coefficient specific for a body region R
(head, chest, abdomen, and so forth.) and for a patient of weight M.
Figure 3 shows (E /
)R,M used in this study to convert energy imparted
obtained from equation 2 into the corresponding value of effective dose; these
values were taken using published data for head
[17] and body
[21,
22] CT examinations.
Differences between the (E / ) for the chest and abdomen are
only about 2%, and the body curves shown in
Figure 3 are the mean chest and
abdomen values.
Protocols
Patient doses in CT examinations depend on the choice of radiographic
technique factors used to perform the scanning. Key parameters that affect
patient dose, and which need to be defined in CT protocols, are the
following:
X-ray tube voltage (kV)—Figure 2 shows how the choice of X-ray tube voltage affects patient dose in CT.
X-ray tube current and scanning rotation time (mAs)—Patient doses will be directly proportional to the choice of mAs.
Axial or helical scanning—In helical scanning, patient doses are inversely proportional to the pitch ratio, assuming a constant scan length. In axial scanning, it is possible to estimate doses by taking the "axial pitch" to be numerically equal to the ratio of the X-ray beam width to the table increment distance, and using this value in equation 1 for determining Dorgan.
Scan length—The total energy put into a patient is directly proportional to the total scan length. For helical scans, one additional rotation (20-mm beam width) was added when computing equation 2 to determine energy imparted to account for the overscanning required for interpolation purposes.
Table 3 presents scanning protocols that explicitly take into account patient size. These protocols are based on those used at our own institution and are generally similar to those adopted in radiology departments that specialize in pediatric imaging (Frush DP, personal communication).
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
30 mGy in newborns to 40
mGy in adults. Figure 6A shows
chest and abdominal CT result in similar doses; it also shows similar trends
with changes in patient weight. In body CT, patients weighing less than
20 kg have organ doses of 7 mGy, which is a factor of 2 lower than
that of a normal-sized (70-kg) adult.
|
|
|
|
Figure 5B shows head CT
effective doses for adults are 0.9 mSv, but that neonate effective doses
are approximately four times higher. Figure
6B shows the effective dose for a neonate undergoing chest CT is
2.2 mSv; the effective dose falls to a minimum of 1.5 mSv for a
10-kg infant and increases to 4 mSv for a normal-sized adult. A neonate
undergoing abdominal CT has an effective dose of 3 mSv; the effective
dose falls to a minimum of 2 mSv for a 10-kg infant and then increases to
3 mSv for a normal-sized adult.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
0.4 seconds or less, and to maintain a reasonable
exposure at the CT detector has required a reduction in the X-ray tube to
isocenter distance. MDCT technology uses X-ray beams that are wider than the
detector array (overbeaming), which also increases patient doses. Further
changes that may increase patient doses include changes in beam-shaping
filters (bow tie) and the use of X-ray tubes with metal frames that have
replaced glass. After the acquisition of new CT equipment, protocols should be
reviewed to help ensure that patient doses are kept as low as reasonably
achievable (ALARA), and that any increase in patient dose is justified by a
corresponding improvement in diagnostic information
[4,
23]. The mean section dose (Dm) shown in Figure 1 relates to an old scanner, whereas those in Figure 4 pertain to an average of major CT scanners in use today. To obtain doses for any current (or future) scanner, it is possible to use the data shown in Figure 1 multiplied by a scale factor equal to the head (or body) CTDI value for the scanner of interest divided by the corresponding HiSpeed Advantage CTDI value given in the footnotes of Table 2. The data shown in Figure 2 show the importance of the choice of X-ray tube voltage (kV) on patient dose. Increasing the X-ray tube voltage from 80 to 140 kV at a constant mAs results in a fivefold increase in patient dose, which is approximately independent of phantom size.
Although most CT examinations to date have been performed at 120 kV, there may be specific diagnostic tasks for which the use of lower kV values could offer substantial dose reductions with no loss of diagnostic information. For example, visualization of iodinated contrast medium administered to patients could be performed at 80 kV to maximize iodine contrast resulting from increased photoelectric absorption by iodine at lower photon energies, which could substantially reduce patient doses [24]. For a novel scanner operating at an X-ray tube voltage that differs from 120 kV, two methods are possible to estimate patient doses at the new kV value [1]: use approximate kV correction factors from Figure 2; or use RkV values measured in head or body dosimetry phantoms.
For head and pediatric CT examinations, the pattern of dose distribution in phantoms and patients is relatively uniform, and peripheral (skin) and central doses are similar. Accordingly, Dm and organ doses for head and pediatric CT examinations may be taken to represent the dose to directly irradiated organs. In larger patients, however, the central dose is expected to be lower than the peripheral dose. Central CTDIs in body phantoms, for example, are generally a factor of 2 lower than those obtained at the periphery [13]. Peripheral-to-central dose ratios as a function of phantom size have been published [25], which enable the average organ doses (Figs. 1, 2, 3, 4) to be used to estimate the expected regional variations in dose for larger patients. CT scanners that use mA modulation will result in a more uniform pattern of dose deposition in patients [26], and mean section doses obtained by use of an average mA are unlikely to be in serious error.
The standard CTDI dosimetry phantoms are made of acrylic and have a
relatively high density of 1.19 g cm-3. The head phantom has a
16-cm diameter, which is equivalent to a water cylinder with a radius of 87
mm, and is therefore similar to water cylinder radii for older children and
heads of adults (Table 1). The
body phantom has a 32-cm diameter, which corresponds to a water phantom with a
radius of 175 mm, and is much larger than typical adult body dimensions
(Table 1). Accordingly, body
CTDI will generally underestimate true patient organ dose because these CTDI
measurements are made in a phantom that is much larger than a normal-sized
adult. Data in Figure 1 show
that correcting for patient size would increase body CTDI by 50% for
abdominal scans and double CTDIs for chest scans. CTDI doses are specified in
acrylic or air, which are typically 10-30% lower than tissue doses
[27]. In addition, CTDI uses
integration distances of either ± seven slice-thickness values or 100
mm, which can underestimate true tissue doses by as much as 70%
[28]. For all these reasons,
CTDIs are poor predictors of patient organ doses; organ doses presented in
Figures 1,
2,
3,
4 will be superior because they
are specified as soft-tissue doses, take into account the patient size, and
have been computed for scan lengths representative of clinical practice
(Table 3).
Organ doses may be used to predict the probability of a deterministic effect such as skin erythema, epilation, and the induction of eye cataracts; a threshold dose of approximately 2 Gy is normally adopted for use in clinical practice [6]. Typical organ doses in head (Fig. 5A, 5B) and body (Fig. 6A, 6B) CT examinations are low (< 0.05 Gy), and deterministic effects are therefore most unlikely for protocols similar to those listed in Table 3. Deterministic effects are nonetheless possible in CT and have been reported for CT perfusion studies, when high techniques are combined with repeated scanning over the same head region [29].
Organ doses presented in this study may be used to estimate embryo or fetal doses in a pregnant patient. Our data suggest that embryo and fetal doses would generally be well below the action threshold of 100 mGy currently recommended by the International Commission on Radiological Protection (ICRP) [30]. Doses presented in Figures 4 and 6A are only approximate, and accurate embryo and fetal dose estimates may need to take into account the pattern of dose distribution in the patient and the location of the embryo or fetus [31].
Note that body organ absorbed doses (7-14 mGy) are considerably lower than those for head CT, whereas body effective doses are generally higher than effective doses for most head CT studies. The effective dose combines organ doses with their relative radiosensitivity (i.e., ICRP weighting factor [4]), and the low head effective dose reflects the absence of radiosensitive organs in the head. Although the chest and abdomen have lower organ doses, they contain the most radiosensitive organs in the body (i.e., lung, red bone marrow, colon, stomach), which results in effective doses that are much higher than those for head CT. Data in Figure 5B show that infants absorb less energy than adults but have effective doses that are about four times higher than those of adults. This increase in effective dose for smaller patients, when scanned at the same techniques as adults, has also been reported by other investigators [32, 33]; high effective doses in infant head CT occur because adjacent organs such as the thyroid and lung are small and their proximity causes them to receive proportionally more scattered radiation [15].
Adult effective doses reported here are somewhat lower than the values of
1.3 mSv recently reported using a similar methodology for head examinations
[15], 5.4 mSv for chest
examinations [16], and
3.9 mSv for abdominal examinations
[14]. Effective doses
generally reflect the choice of scanning protocol used to perform a given CT
examination, and our computed effective dose values are low because we used
low mAs values, minimized the scan length, and used a pitch of 1.5 for helical
scanning (Table 3). It is
noteworthy that current pediatric doses are now much lower than reported
previously; for a 10-kg patient, the effective dose for a chest CT examination
has been reduced from 9.6 to 1.5 mSv
[16], and the corresponding
reduction for an abdominal CT examination is from 7 to 2 mSv
[14]. Because of the increased
radiosensitivity of children and the increasing utilization of CT in this age
group, it is most welcome that pediatric doses can be significantly reduced
with no apparent loss of diagnostic information
[1,
34].
Our methodology is based on modeling patients as uniform cylinders of water for the purposes of estimating the pattern of energy deposition and the total energy imparted to the patient. This simplification should not result in large dose errors because the attenuation properties of water for CT spectra, where the Compton effect is dominant, would be similar to those of patients, who are composed of bone and tissue. Our approach should approximate absolute patient dose values, but is expected to provide reliable relative dose estimates when technique factors (kV, mAs, pitch) or the CT scanner type are changed.
One major limitation of our study is that we have not explicitly included the effect of X-ray tube mA modulation as the X-ray tube rotates around the patient or along the patient's long axis. For a given body region, patient doses are expected to be directly proportional to the average mAs value. If the average mAs values were lower for a given CT scanner than those used in this study (Table 3), then the patient doses would be correspondingly lower.
Patient radiation risks may be estimated from the effective doses given in this study. For example, the ICRP specifies that the nominal risk coefficient for the induction of fatal cancer is 5%/Sv, and the total detriment (i.e., induction of all cancers and genetic effects) is 7.3%/Sv when risk factors are averaged over a whole population [4]. However, note that any risk estimates need to take into account patient demographics, with large differences in the risks per unit effective dose between infants and older adults [8]. Of greater importance is the fact that there are large uncertainties in risk estimates at dose levels normally encountered in CT [2]. One helpful way to better understand the significance of a given CT effective dose value is to compare it with other types of radiation exposure encountered in society at large (Table 4). Patient doses in CT are broadly comparable to annual doses from natural background, and any risks from a single CT scan should be comparable to those from background doses delivered to every inhabitant of our planet each year.
|
Acknowledgments
We gratefully acknowledge the technical support provided by M. L. Roskopf
and R. L. Lavallee and valuable discussions with D. P. Frush and T. Toth.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  F. A. Mettler Jr, W. Huda, T. T. Yoshizumi, and M. Mahesh Effective Doses in Radiology and Diagnostic Nuclear Medicine: A Catalog Radiology, July 1, 2008; 248(1): 254 - 263. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. J. Francois, D. Tuite, V. Deshpande, R. Jerecic, P. Weale, and J. C. Carr Unenhanced MR Angiography of the Thoracic Aorta: Initial Clinical Evaluation Am. J. Roentgenol., April 1, 2008; 190(4): 902 - 906. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. R. Virta, T. Tolvanen, K. Nagren, A. Bruck, A. Roivainen, and J. O. Rinne 1-11C-Methyl-4-Piperidinyl-N-Butyrate Radiation Dosimetry in Humans by Dynamic Organ-Specific Evaluation J. Nucl. Med., March 1, 2008; 49(3): 347 - 353. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Huda Radiation Doses and Risks in Chest Computed Tomography Examinations Proceedings of the ATS, August 1, 2007; 4(4): 316 - 320. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. P. Forman Back to the Beginning Am. J. Roentgenol., February 1, 2007; 188(2): 295 - 296. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
