|
|
||||||||
Original Research |
1 All authors: Department of Radiology, Duke University Medical Center, Erwin Rd., Box 3808, Durham, NC 27710.
Received July 23, 2007;
accepted after revision November 9, 2007.
R. C. Nelson is a consultant for GE Healthcare.
Abstract
|
|
|---|
MATERIALS AND METHODS. Estimated absorbed fetal dose was measured using a metal oxide semiconductor field effect transistor (MOSFET) dosimeter that was placed in the uterus of an adult female anthropomorphic phantom. The phantom was scanned on a 16-MDCT scanner using three protocols. The scanning parameters for protocol A (trauma) were detector configuration, 16 x 0.625 mm; pitch, 1.75:1; rotation time, 0.5 second; 140 kVp; and 340 mA. The scanning parameters for protocol B (CT angiography) were detector configuration, 16 x 1.25 mm; pitch, 1.38:1; rotation time, 0.6 second; 140 kVp; and 300 mA. The scanning parameters for protocol C, which is the automated tube current modulation (ATCM) protocol previously used in the literature, were detector configuration, 16 x 1.25 mm; pitch, 0.938:1; rotation time, 0.5 second; 140 kVp; and 380 mA. The protocols were also modified for the ATCM mode; the CTDIvol was documented from the scanner's console. Correlation between these data was tested with a goodness-of-fit model.
RESULTS. Absorbed fetal radiation dose in the early first trimester correlated with the CTDIvol via a linear regression equation. For a constant tube current and peak voltage of 140 kVp, fetal dose (mGy) = 1.665 x CTDIvol (mGy) – 7.059. For the ATCM mode and a constant kVp of 140, fetal dose (mGy) = 2.151 x CTDIvol (mGy) – 2.200. The goodness of fit (R2) for the equations is 0.99 and 0.91, respectively.
CONCLUSION. In both the manual and ATCM modes, absorbed fetal radiation dose can be estimated from the CTDIvol obtained at the time of scanning independent of pitch and tube current–time product (mAs).
Keywords: automated tube current modulation mode fetal imaging MDCT radiation dose
|
|
|---|
Fetal dose estimation for axial and helical CT scanners has been described in the literature [7, 19–21]. Subsequent studies used methods to estimate fetal dose from CT examinations in late pregnancy based on normalized weighted dose indexes and peak voltage and tube current settings or by using phantom data and the CT dose index (CTDI); the latter studies used thermoluminescent detectors and phantom calculations [7, 19, 20]. Other investigators have used computerized modeling (e.g., the Monte Carlo geometric model) and measurable markers such as skin entrance dose or CTDI to predict fetal dose [7, 22–24]. Hurwitz et al. [15] and Yoshizumi et al. [25] measured fetal doses for MDCT body protocols with an anthropomorphic phantom using metal oxide semiconductor field effect transistor (MOSFET) detectors placed in the location of the uterus; these data are a more physiologic simulation of fetal dose than computerized modeling or skin entrance dose.
To date, to our knowledge, there has been no study of the direct correlation between absorbed fetal dose and scanning data (CTDIs) as a method to estimate fetal dose. Because volume CTDI (CTDIvol) reflects pitch-adjusted dose in the x-, y-, and z-planes and is not indicative of patient height or scan length, this parameter was chosen to reflect a scanner-measured dose. Although studies have shown that the application of the automated tube current modulation (ATCM) mode decreases radiation dose [26–30], to our knowledge no method has been proposed for the fetus dose (or any other organ dose) estimation with the ATCM mode in modern MDCT scanners. The purpose of this study was to determine the relationship between recorded CTDIvol of a maternal scan of the chest, abdomen, and pelvis on a 16-MDCT scanner and directly measured absorbed fetal dose using both the manual and ATCM modes.
|
|
|---|
|
|
|
Scanning Parameters
The anthropomorphic phantom was imaged on a 16-MDCT scanner (LightSpeed 16)
using current clinical protocols (Fig. 1C). The scanning parameters for the
following pro tocols are listed in Table
1: trauma (protocol A); CT angiography (protocol B); and scanning
of the chest, abdomen, and pelvis with the parameters previously used in the
literature (protocol C) [26].
The anatomic landmarks for scanning were to begin at the thoracic inlet and
terminate at the pubic symphysis. Each protocol was scanned at a fixed tube
current (manual mA mode) and using the manufacturer's ATCM mode (Smart mA, GE
Healthcare). When the ATCM mode is used with a GE scanner, the noise index
(GE-specific selection) and maximum and minimum tube current thresholds must
be selected. We used noise indexes of both 10 and 12.5 H, values that are
recommended by the manufacturer for standard abdominal CT and those reported
in the literature to show acceptable image noise while reducing radiation
exposure
[27–30].
The phantom was scanned three times for each of the nine protocols, and the
average absorbed dose and SD were calculated.
|
Correlation with CTDI and Statistical Measurements
For each scan, the CTDIvol was recorded from the CT scanner's
operator console. We measured the relationship between documented
CTDIvol and aver age fetal radiation dose and goodness of fit for
this model with a statistical package (Prism 4, GraphPad Software). Dose
comparisons and com parisons of CTDIvol values were made from a
two-way analysis-of-variance test (SAS 9.1.3, SAS Institute).
|
|
|---|
|
A linear relationship between the absorbed fetal dose and
CTDIvol values was determined (Figs.
3A and
3B). The equation for the
manual tube current (mA) mode is as follows:
![]() |
|
|
![]() |
|
|
|---|
Our study corroborates several early estimates of fetal dose from MDCT found in the literature [8, 12, 15, 19, 23]; these values are all below the reported dose threshold for the induction of teratogenesis [10, 14, 16]. As we anticipated, the estimated fetal doses significantly decreased with the use of the ATCM mode; this result corroborates the results of past studies of this technique [26, 27, 30–32]. The results of our study support the hypothesis that given a constant kVp, the correlation between CTDIvol and absorbed fetal dose in early pregnancy follows a linear relationship; this relationship was expected because radiation dose is known to vary linearly with changes in tube current. For each vendor-specific approach to ATCM, the manufacturer has enabled operator-selected scanning with choices that trade image noise and reconstruction slice thickness for decreased patient dose. We found that a direct correlation between absorbed dose and CTDIvol values exists for the ATCM mode of imaging as well, likely because tube current stays relatively constant during scanning of the pelvis in order to maintain imaging technique in the presence of the surrounding pelvic bones [33]. The linear correlation of the CTDIvol and absorbed fetal dose in ATCM mode may also be related to the location of the uterus relatively centrally within the pelvis so that variations in dose based on relationship to the X-ray beam are not as apparent. We are now able to prove that despite several variables (e.g., human phantom body shape is not symmetric, effect of scatter dose on fetus is unknown), reported CTDIvol does show a linear relationship with absorbed organ dose. The conversion factor in our equation presents the best fit to the raw data under the manual and ATCM modes.
Our protocols were selected to simulate the possible scanning parameters used in routine clinical practice. These protocols are those used by emergency departments in everyday patient care. We have included the entire thorax in these protocols to reflect routine trauma scanning. We have also included an angiographic protocol because the scanning parameters differ from those of a routine trauma study (detector configuration, 16 x 1.25 vs 16 x 0.625 mm, respectively); this protocol includes the thorax as well. The impact of thoracic imaging on fetal dose is well known; only scatter radiation is involved and the dose to the fetus is small, with reports ranging between 20 and 320 µGy, depending on the scanning parameters [15, 34]. These doses add little to the overall fetal dose and should be considered negligible.
Our equations to estimate fetal dose are dependent on knowledge of CTDIvol for the scan in question. In the manual mode, tube current is uniform and thus CTDIvol has little variation along the z-axis. The ATCM mode, in contrast, varies current along the length of the scan, and CTDIvol is calculated from the average current throughout the entire scan regardless of scanner vendor or type of ATCM used. Our equation for the ATCM mode represents an averaged CTDIvol value for the scanning protocol and underestimates the effect current changeability will have on fetal dose. Note that as overall patient size increases, radiation dose will increase because scanning will require currents greater than the upper limits of the ATCM mode setting. Given our phantom's body habitus, it is likely that our calculation closely approximates absorbed dose in the ATCM mode, despite expectations of variability in current. For large patients, current variability and total increase in dose may more significantly impact overall dose and dose to the fetus. Further research in this area will prove helpful.
The results of our study have implications beyond determining fetal radiation dose in early gestation. To our knowledge, this study is the first reported correlation between CTDIvol and calculated absorbed organ dose. Extrapolation of this index for use in determining individual organ absorbed dose may extend its applicability into areas of radiation protection and MDCT protocol and scanner design. The results from this study give the first insight into how this index may be used in these ways. We suspect that irradiation in the first trimester may be a cause for alarm in the general population [35–37], and this method may be helpful for the radiation biologist to use in counseling the referring physician, so the mother is not forced to terminate her pregnancy.
There are limitations to our study. First, data have shown that fetal depth affects absorbed dose and that fetal depth is dependent on multiple variables, including stage of pregnancy, uterine position, and bladder volume [9, 19, 38]. It is possible that there is a range of fetal doses based on variations in uterine positions in early pregnancy, as has been noted in the second and third trimesters [38]. We did not vary uterine position in our study.
Second, two studies have shown that bladder volume significantly affects fetal depth in the first trimester, with a decrease in fetal depth seen with an empty bladder [9, 39]. There are, however, no measurable data to suggest that these variations would considerably impact estimated dose from MDCT in the early first trimester.
Third, maternal body mass index has also been shown to be related to fetal depth and thus fetal dose [38, 39]. We used a phantom with a small body habitus. We were therefore unable to examine the contribution of maternal size to absorbed fetal dose. Extrapolation from other studies suggests that in patients with increased body mass index, the absorbed dose for internal organs is lower for constant CT parameters [40]. The effects in the ATCM mode suggest that dose is, in fact, higher because of the current needed to overcome variations in noise [28]. Similarly, this study was designed to address potential fetal radiation in the first trimester; the physiologic and morphologic changes seen in the second and third trimesters, including but not limited to the effect of uterine ascent in the abdomen, decreasing amounts of amniotic fluid, and changes in the maternal body habitus, are not addressed in this study. Further investigation with modified phantoms will be revealing.
A final limitation of our study, and perhaps the most substantial one, is that our equations were calculated using data from a single manufacturer's scanner and a set kVp and were performed only on a 16-MDCT scanner. The data for this study were acquired at a time when 64-channel scanners were not available. Since there may, in fact, be some variability among vendors' scanners and detector configurations, further research will establish whether this relationship varies for different scanner configurations and methods of dose modulation.
In conclusion, absorbed fetal radiation doses are lower with the ATCM mode than with the manual mA mode during early pregnancy. Given a constant kVp, there is a linear relationship between fetal dose and CTDIvol in both the manual and ATCM scanning modes that allows one to estimate fetal dose from a single-variable equation. This approach can potentially be applied universally regardless of the pitch and other imaging parameters.
|
|
|---|
This article has been cited by other articles:
![]() |
S. Matsuoka, A. R. Hunsaker, R. R. Gill, I. B. Oliva, B. Trotman-Dickenson, F. L. Jacobson, and H. Hatabu Vascular Enhancement and Image Quality of MDCT Pulmonary Angiography in 400 Cases: Comparison of Standard and Low Kilovoltage Settings Am. J. Roentgenol., June 1, 2009; 192(6): 1651 - 1656. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Matsubara, K. Koshida, K. Ichikawa, M. Suzuki, T. Takata, T. Yamamoto, and O. Matsui Misoperation of CT Automatic Tube Current Modulation Systems with Inappropriate Patient Centering: Phantom Studies Am. J. Roentgenol., April 1, 2009; 192(4): 862 - 865. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. Tack and P. A. Gevenois Body MDCT at 140 kV Am. J. Roentgenol., March 1, 2009; 192(3): W139 - W140. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |