| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Division of Pediatric Radiology, 1905
McGovern-Davison Children's Health Center, Box 3803, Department of Radiology,
Duke University Medical Center, Durham, NC 27710.
2 Division of Radiation Safety, Duke University Medical Center, Durham,
NC.
3 Department of Radiology, Stanford University Medical Center, Palo Alto,
CA.
Received November 13, 2006;
accepted after revision February 19, 2007.
Address correspondence to C. L. Hollingsworth
(holli016{at}mc.duke.edu).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. ECG-gated cardiac CTA simulating scanning of the heart was performed on an anthropomorphic phantom of a 5-year-old child on a 16-MDCT scanner using variable parameters (small field of view; 16 x 0.625 mm configuration; 0.5-second gantry cycle time; 0.275 pitch; 120 kVp at 110, 220, and 330 mA; and 80 kVp at 385 mA). Metal oxide semiconductor field effect transistor (MOSFET) technology measured 20 organ doses. Effective dose calculated using the dose–length product (DLP) was compared with effective dose determined from measured absorbed organ doses.
RESULTS. Highest organ doses included breast (3.5–12.6 cGy), lung (3.3–12.1 cGy), and bone marrow (1.7–7.6 cGy). The 80 kVp/385 mA examination produced lower radiation doses to all organs than the 120 kVp/220 mA examination. MOSFET effective doses (± SD) were as follows: 110 mA: 7.4 mSv (± 0.6 mSv), 220 mA: 17.2 mSv (± 0.3 mSv), 330 mA: 25.7 mSv (± 0.3 mSv), 80 kVp/385 mA: 10.6 mSv (± 0.2 mSv). DLP effective doses for diagnostic runs were as follows: 110 mA: 8.7 mSv, 220 mA: 19 mSv, 330 mA: 28 mSv, 80 kVp/385 mA: 12 mSv. DLP effective doses exceeded MOSFET effective doses by 9.7–17.2%.
CONCLUSION. Radiation doses for a 5-year-old during cardiac-gated CTA vary greatly depending on parameters. Organ doses can be high; the effective dose may reach 28.4 mSv. Further work, including determination of size-appropriate mA and image quality, is important before routine use of this technique in children.
Keywords: CT angiography • dosimetry • pediatric radiology • radiation dose
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
500 milliseconds) narrower (submillimeter) detector widths, and
increasing numbers (e.g., 64) of detector rows combining to improve image
quality by increasing spatial resolution and decreasing motion artifacts. In
addition, image degradation from cardiac motion is further reduced using
retrospective ECG gating or prospective ECG triggering
[6–8]. Cardiac-gated CTA is being used in the assessment of coronary arteries in adults [9–12]. However, imaging the heart and coronary arteries using cardiac-gated CTA has not been systematically assessed in children. Systematic assessment includes determination of the parameters that contribute to radiation dose from cardiac-gated CTA. For example, optimal tube current and kilovoltage settings are still evolving in adults even as the number of cardiac-gated CTA examinations is increasing [13]. In addition, attention has recently been directed to dosimetry of cardiac-gated CTA [14–23]. However, one difficulty in assessing CT dose is that traditional methodology using thermoluminescent dosimeters (TLDs) is problematic, especially for assessing multiple CT examinations because multiple phantoms or repeated loading of dosimeters into a single phantom is necessary. An alternative method of dose assessment using the dose–length product (DLP) has been studied but has recently been challenged as imprecise [24, 25]. A relatively new combination of dosimetry using metal oxide semiconductor field effect transistor (MOSFET) technology and adult and pediatric anthropomorphic phantoms has overcome many of these technical dosimetry difficulties [26–35]. To our knowledge, to date this combination has not been applied to pediatric cardiac-gated CTA radiation dose assessment.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MOSFET detectors were calibrated as follows. First, we determined the thickness of copper sheets to achieve the half-value layer (HVL) (7.24 mm Al at 120 kVp) of GE Healthcare CT scanners [27] with a conventional radiographic X-ray tube; we added 0.2-mm copper sheets to the X-ray tube to obtain an equivalent HVL of 7.37 mm Al at 120 kVp. After this had been accomplished, individual MOSFET detectors were calibrated at a clinical energy of 120 kVp. Individual calibration factors were obtained for all 20 MOSFET detectors by fitting four data points with the least-squares fit routine (Prism, version 2.0, 1995, GraphPad software). These conversion factors were stored in the MOSFET software (AutoSense PC software version 2–1, TN-RD-49, Thomson-Nielsen) for immediate readout after each protocol was performed. The bias supplies provide a regulated bias voltage to the detectors and are connected to a reader after a radiation exposure to measure the threshold voltage shift in the detector. This permanent shift in the threshold voltage after irradiation is proportional to the absorbed radiation dose [38].
The MOSFET sensors gathered data from 20 simulated anatomic sites in the tissue-equivalent phantom (Table 1). The leads were placed to measure doses at specific organ locations in the phantom, which was configured of individual slices containing anthropomorphic tissue equivalents (Fig. 1). The data were directly sent to a laptop computer after each exposure. Subsequently, the effective radiation dose was calculated according to guidelines published in International Commission on Radiological Protection (ICRP) 60 by summing the products of the average recorded organ radiation dose and the ICRP weighting factor [39]. Radiation doses measured in tissue equivalents for small organs such as the ovary and thyroid gland were through a single-detector scanner only. In comparison, radiation doses measured in larger organs such as the liver or lung were calculated after determination of the mean of several-detector dose measurements. Pediatric bone marrow distributions were previously published by Cristy [40]. We have used percentage of bone marrow distribution data for 5-year-olds in the effective dose calculations for the major skeletal regions: skull (cranium + mandible) (17.44%), ribs (ribs + sternum) (10.58%), spine (middle portion) (9.58%), and pelvis (23.33%).
|
The effective dose was also calculated by the DLP displayed on the CT
console for each protocol. The general formula for effective dose is as
follows:
 |
The retrospective gated cardiac CT examinations were performed on a 16-MDCT scanner (LightSpeed, GE Healthcare) using the SnapShot Burst Plus software (GE Healthcare), which creates an image made with 180° rotation. Multisector reconstruction is used to improve temporal resolution to better than 250 milliseconds. The specific scanning parameters are summarized in Table 2. In the absence of published protocols for pediatric cardiac-gated CT angiography, three different protocols were arbitrarily designed to cover a range of potential doses delivered during gated CT angiography. These diagnostic phase protocols consist of modifications of tube current (in milliamperes) arbitrarily designated as high (330 mA), medium (220 mA), and low (110 mA), corresponding to the milliampere value. ECG dose modulation was not available, and thus mA was fixed throughout each examination. The peak kilovoltage (kVp) was kept constant at 120 kVp for these tube currents. The high mA protocol was chosen to represent the tube current that is currently in use in clinical practice in adults because conceivably that could be a default protocol for pediatric imaging [42]. We were also interested in assessing dose for a lower kVp protocol because reduced kVp has been advocated in pediatric MDCT, including CTA [35], so an additional protocol using 80 kVp and an mA of 385 was performed. Based on data that a decrease in kVp from 120 to 80 results in an increase in noise of 68% [43], an increased tube current was used to partially compensate for this. The tube current was increased from 220 to 385 mA (75% increase), yielding a predicted decrease in noise of 33%. Because contrast improves with lower kVp, a balance in noise was not attempted. This protocol was not selected to be an equivalent in noise to the 220 mA/120 kVp protocol, but is a representation of a protocol used at one investigator's institution after consultation with the manufacturer of the MDCT scanner. Dosimetry was also performed for the timing bolus exposure at 5-mm axial acquisition, 0.8 second gantry rotation time, 40 mA, and 120 kVp.
|
The phantom was examined in the supine position with a single diagnostic phase examination from the carina to the cardiac apex (10 cm) using retrospective cardiac gating (Fig. 2). A heart rate simulator (ECG Simulator, model GE Marquette, GE Healthcare) was used, set at a rate of 100 beats per minute representing a normal heart rate for a 5-year-old child [44]. Each of the four diagnostic phase examinations and a single timing bolus phase were performed three consecutive times. The organ doses obtained during each of the three exposures were used to obtain an average dose value and SD for each protocol. During each examination, the same dosimeters were exposed in identical locations to minimize recorded dose bias due to intrinsic measurement disparities. The effective organ doses discussed in this article are therefore the summed averages of the three simulated scans for each protocol. MOSFET effective dose was determined using ICRP 60 guidelines [45, 46] and compared with the DLP effective dose, calculated using methodology by Shrimpton and Wall [41].
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
Organ dose measurements varied depending on the specific protocol evaluated. When assessing the doses calculated for the three protocols with constant kVp (120 kVp), the low-mA (110 mA) protocol imparted a lower radiation dose to all organs than the medium-mA (220 mA) or the high-mA (330 mA) protocols, as expected. However, this investigation provided an opportunity to assess organ doses (Table 1). We found that the highest organ dose from all protocols was to the lungs and the breast. The combined breast dose was 3.4 cGy (right and left breasts averaged) for the low-mA (110 mA, 120 kVp) protocol, 8.0 cGy for the medium-mA (220 mA, 120 kVp) protocol, and 12 cGy for the high-mA (330 mA, 120 kVp) protocol. The lung dose (top, middle, and lower lung lobe doses averaged) was 3.5 cGy for the low-mA (110 mA, 120 kVp) protocol, 7.5 cGy for the medium-mA (220 mA, 120 kVp) protocol, and 11.6 cGy for the high-mA (330 mA, 120 kVp) protocol. The 385 mA/80 kVp protocol delivered 5.3 cGy as a breast dose and 4.6 cGy as a lung dose, which are lower doses than either the medium-mA or high-mA protocols.
The MOSFET effective doses for the four different protocols are provided in Table 3 and Figure 4. These results show the expected reduction in the resultant effective dose as the mA is reduced by one third for the medium- and by two thirds for the low-dose protocols, as compared with the high-mA protocol. That is, our results show the predicted drop in effective dose from 28.4 to 19.9 mSv (measured, 30%; expected, 33.3%) and finally to 10.1 mSv (measured, 64%; expected, 66.7%) as the tube current changed from 330 to 220 mA and finally to 110 mA.
The effective dose for the 80-kVp examination, including the timing bolus, was 13.3 ± 0.2 mSv. The low-kVp examination as designed (385 mA, 80 kVp) had a lower effective dose than the CT protocols with medium mA (220 mA, 120 kVp) and high mA (330 mA, 120 kVp). For diagnostic scanning, the effective dose calculated from the DLP method was always higher than the measured effective dose, and differed from between 9.7% to 17.2% (Fig. 5).
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Therefore, establishing MDCT radiation doses is a necessity for both clinical management of patients and for institutional review board risk assessment for research. However, establishing these doses has been problematic. The traditional method for quantification of radiation dose using thermoluminescent dosimeters (TLDs) is time-consuming. For example, loading, unloading, and processing the TLDs must follow each CT examination. Therefore, only one CT examination can be performed at a time (unless another phantom with TLDs is available), and dosimeters may have to be shipped to another site for interpretation. The MOSFET technique for dosimetry provides real-time dosimetry in which data from multiple CT examinations can be acquired instantaneously during a single session. MOSFET dosimetry is particularly useful in the comparison of doses from multiple CT protocols in which the dose effect of various manipulations such as alteration of mA, kVp, or pitch can be assessed. MOSFET technology has previously been validated against the TLD method; it also provides an accurate method to assess dose, with uncertainty in dose recorded during the three separate runs for each protocol in the range of 10% at the 1-cGy dose level [25, 34].
Based on this MOSFET dosimetry, cardiac-gated CTA doses to a 5-year-old can exceed 28 mSv when a protocol with adult techniques is used. These doses are well beyond those used for routine chest CT in adults (5.4 mSv) or children (< 5 mSv) [25, 54, 55]. For the child, the dose is up to 5.6 times the radiation dose of routine chest CT. For diagnostic scanning, the effective dose calculated using the DLP method was consistently higher (9.7–17.2%) than the effective dose calculated from the ICRP 60 guidelines [39]. The DLP is available on most modern CT scanners, and our data show that although overestimation occurs consistently, the DLP method provides a reasonable estimate, we believe, for the effective dose for cardiac-gated CTA in a 5-year-old. This difference, however, may not be reflected at different ages (cross-sectional areas) or when using a different field of view, and it obviously depends on the conversion factor used.
Our dosimetry data also provide an opportunity to compare radiation doses from cardiac-gated CTA with those from conventional angiography. The radiation dose delivered to the 5 year-old using an adult protocol also exceeds the dose reported for adults. Several recent investigations of the radiation dose delivered by coronary cardiac-gated CTA in adults have found the potential radiation dose to be relatively high (up to 18.8 mSv) depending on the type of scanner [21, 23, 56]. Using results from Hunold et al. [13], our data also indicate that the dose delivered by cardiac-gated CTA may be as great as 2.5 times the dose delivered by conventional angiography. Bacher et al. [57] report median effective doses of 4.6 mSv for diagnostic pediatric cardiac catheterizations and 6.0 mSv for therapeutic cardiac catheterizations. Similar measurements are reported by Rassow et al. [58], who found effective dose measurements from cardiac catheterization in infants to range from 2 mSv (25th percentile) to nearly 18 mSv (90th percentile). In comparison, the preliminary data with our pediatric phantom indicate that there is potential for the radiation dose to children undergoing cardiac-gated CTA to be substantially higher than the dose delivered by standard cardiac catheterization. If a high-dose technique is used, the dose to a child for cardiac-gated CTA may be up to six times (28 vs 4.6 mSv) that of a standard cardiac catheterization.
As shown in an anthropomorphic phantom of a 5-year-old, cardiac-gated CTA in children can result in a higher radiation dose than conventional angiography. However there are benefits of CT versus conventional catheterization in children. Cardiac catheterization is invasive. Complications associated with femoral artery catheterization include occlusion, dissection, pseudoaneurysm formation, and retroperitoneal hemorrhage [57, 58]. Furthermore, serious sequelae such as limb growth discrepancy can occur because of arterial ischemia in infants and young children [55, 58–61]. Postprocedural monitoring may require several hours in a recovery unit. Cardiac catheterization may also require general anesthesia. Cardiac-gated CTA does not require the same frequency of sedation as conventional angiography and requires only that there be adequate venous access [35, 62]. Although no reports of sedation frequency are available to our knowledge, the limited need for sedation in these examinations would likely also apply because the examination is rapidly acquired and can be performed in seconds. Cardiac-gated CTA can easily be completed using 1.5 mL/kg of IV contrast material, which is less than the amount generally required for cardiac catheter angiography [35]. If the IV access site allows power injection, no ancillary staff is exposed to radiation (e.g., when a manual bolus technique is used, the individual responsible for the injection must be in the room during at least a portion of the image acquisition). The procedure itself imparts a radiation dose to the child and the physician performing the examination, and any support staff who are required to be present in the angiography suite [63–65].
Because CT angiography in children can be performed at a lower kVp, which will increase the inherent tissue contrast [35, 43, 63] and potentially improve the contrast-to-noise ratio, we elected to study a lower-kVp protocol. It was not our intention to maintain image noise compared with the three protocols performed at 120 kVp, but to select a kVp level that has been advocated in the literature for pediatric CTA and to partially compensate for the increased noise with an increase in tube current to 385 mA. With this protocol, the measured dose was less than with the 220 mA/120 kVp protocol but higher than with the lowest mA protocol. These data indicate that dose levels can be achieved in the lower range when a lower kVp is used, although more work must be done to assess the overall affect on image quality, considering both mottle and contrast.
Our investigation has several limitations. The examination was performed for only one age phantom. We chose a phantom of a 5-year-old because of availability and because this phantom better approximated the range of children evaluated from infancy through early teens. In addition, the examinations were all performed on a single scanner type with select protocols. Although the adult parameters arguably would not be the default, to our knowledge no pediatric cardiac-gated CTA guidelines have been systematically investigated. We chose the specific protocols in this investigation to represent high-, medium-, and low-dose protocols to assess a range of potential doses.
Another limitation is that there was no investigation of image quality or the potential effect of changes in protocol on image quality. Our investigation also did not evaluate the potential benefits (decreased dose) due to dose modulation techniques because our institution did not have this capability during the period of the investigation. It has been reported that this technique can reduce the radiation dose by 30–50% by significantly reducing the tube current during systole [66]. Also, some discrepancies were seen in organ doses measured by MOSFET. We believe the discrepancy in dose to the breast relative to that imparted on the lungs with the low-mA (110 mA) protocol may be explained by the helical rotation of the X-ray tube and the variation of the dose measured when the breast is positioned as a surface organ versus its exposure when the X-ray beam passes through the chest from back to front. The lungs received a more uniform exposure because of their relatively central position in the chest. The changes in relative position of the breast and the X-ray beam source may also explain slight differences in dose received by the right and left breasts. We performed three runs for each protocol to reduce the positional effect of the detectors. Although performing an axial acquisition of data would eliminate this sampling issue based on detector position, this technique is not practical for cardiac-gated evaluation.
Finally, we used an estimated conversion factor of 0.021 mSv ·
mGy–1 · cm–1 for pediatric chest MDCT
based on justifications in the methodology. In fact, when we computed a dose
conversion factor (effectiveDLP) from MOSFET results, we arrived at
a conversion factor of 0.019 using the equation:
 |
In conclusion, technical advances in cardiac imaging with cardiac-gated CTA are rapidly evolving. MOSFET technology is a useful tool for measuring radiation dose from cardiac-gated CTA protocols and can provide radiation dose quantification. Depending on parameters, these doses vary substantially and can be relatively high, with the effective dose reaching 28.4 mSv in the protocols tested; some of the highest organ doses are to the pediatric breast. More information is needed for development of optimal scanning parameters and assessment of the diagnostic accuracy of this technique before its widespread use in the pediatric population.
|
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. Coursey, D. P. Frush, T. Yoshizumi, G. Toncheva, G. Nguyen, and S. B. Greenberg Pediatric Chest MDCT Using Tube Current Modulation: Effect on Radiation Dose with Breast Shielding Am. J. Roentgenol., January 1, 2008; 190(1): W54 - W61. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
