| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Full paper |
1 Academic Unit of Medical Physics, University of Leeds, 2 Department of Medical Physics and Engineering, Leeds Teaching Hospitals NHS Trust, Leeds General Infirmary, Leeds LS1 3EX, UK
Correspondence: Mr Gareth R Iball, Medical Physics & Engineering, Leeds Teaching Hospitals NHS Trust, Wellcome Wing, Leeds General Infirmary, Leeds LS1 3EX, UK. E-mail: gri{at}medphysics.leeds.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
The British Thoracic Society guidelines state that computerized tomography pulmonary angiography (CTPA) is currently the recommended imaging modality for the diagnosis of PE, although ventilation/perfusion scanning is often used [9]. These guidelines do not include specific recommendations for pregnant patients.
The radiation dose to the fetus from CT scans aiming to diagnose PE has been estimated in a number of studies [1, 10–13]. In a study using anthropomorphic phantoms. Hurwitz et al [10] report a fetal dose of between 0.2 and 0.7 mGy. Fetal doses have been estimated using mathematical models [1], and doses of 0.003–0.130 mGy and 0.14 mGy have been reported [1, 11]. The fetal dose will increase during pregnancy both as the fetus grows in size and as it moves closer to the scan volume [1].
The developing fetus is sensitive to radiation because of the highly proliferating nature of the cells and processes of differentiation [14]. The risks of causing a range of different radiation effects vary with gestational age [15]. The expected doses involved in CT are not likely to meet the threshold for inducing malformations or reduction in intelligence, so the most important consideration is the induction of childhood cancer. It is likely that this risk is constant over the entire term, and there is no evidence that the pattern of cancer induction varies with gestational age [14], or that the induction of cancer has a dose threshold. The excess risk of the induction of childhood cancer is reported as 1 in 33 000 per mGy [15].
The use of lead shielding
The use of lead shielding to provide protection from irradiation in CT has been well reported for a wide range of organs [13, 16–24]. In this work, we consider the case where the organ at risk, the fetus, is subject only to internally and externally scattered radiation. This means that the detrimental effects on image quality of placing the shielding within the primary beam are avoided.
The use of lead shielding for the protection of the uterus and the ovaries in CT scanning of the upper abdomen is considered by Hidajat et al [16]. No reduction in the dose to the uterus and ovaries was found through the use of the lead shield, although the authors make two important points: first, that (from their own observation) the internal dose is always greater than the surface dose for measurements outside the primary beam and, second, that the effectiveness of the lead shielding is dependent on whether the organ to be protected lies superficially or deep in the body.
Doshi et al [13], in their work measuring fetal doses during CTPA scans, report a reduction of approximately 30% when using a lead apron as shielding. The phantom used in their study was based on a pregnant patient at full term.
Dose reductions were observed in three studies considering the radiation dose to the male gonads using a 1 mm lead shield designed to surround the testes [16, 17, 22]. Price et al [22] report dose reductions of 82% where the testes were 15 mm from the edge of the scan field and 77% where the testes were 10 mm from the edge of the scan field during pelvic CT scanning in a situation where the dose is only due to scatter. Hohl et al [17] found that the dose to the testes was reduced by 87% during routine abdominopelvic CT scanning, and Hidajat et al [16] used a phantom study to demonstrate a 95% reduction in the dose to the testes in abdominal CT examinations.
Shielding of the breasts during CT scanning of the head has been discussed by two studies [21, 23]. Beaconsfield et al [23] showed that the dose to the breast was reduced from 0.32 mGy to 0.042 mGy by the use of a lead shield, which represents a reduction of 87%. Brnic et al [21] also reported a reduction in breast dose of 57% through the use of a 0.35 mm lead shield. Measurements were also made by Brnic et al [21] to analyse the relative contribution from internal and external scatter. The results of these measurements imply that the relative contributions are approximately equal and that the ratio of internal to external scatter may be dependent on the depth of the measurements made.
Although no dose reduction was reported by Hidajat et al [16], the above, along with the work of Doshi et al [13], suggests that a dose reduction should be possible.
Aims of this work
This work aims to determine whether lead shielding can be used to decrease the radiation dose to the fetus during CT scans for the diagnosis of PE during early stage pregnancy. This information can be used to produce a protocol that will yield the minimum practically achievable dose to the fetus while preserving image quality within the scan volume.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Designing the phantom
A Rando phantom (Alderson Laboratories, Stamford, CT) was modified to enable the dose to the fetus to be measured directly (Figure 1a
). Information reported in Adams et al [12] was used to determine the distance of the top of the uterus from the top of the head. To verify the uterus position, the bones that were visible in the phantom were also matched to anatomical images [25]. These two methods indicated slice number 31, which had also been used by Hidajat et al [16], as the optimal position for the uterus. The organ dose to the uterus was taken as being equal to the fetal dose [15]. A Perspex replacement for slice 31 was machined with a well so that a 15 cc ionization chamber and the attached cables could be fitted (Figure 1b). Perspex is known to have a higher linear attenuation coefficient than the Rando phantom muscle material of the order of 10% at diagnostic energies and, as such, the replacement slice may have attenuated the scattered radiation more than the slice of the phantom that was removed. This study is designed to evaluate the relative change in fetal dose when lead shielding is used and, as such, the difference in attenuation between Perspex and the phantom material is thought to have had a minimal effect on the results.
|
Experimental set-up and protocol
A Siemens SOMATOM Sensation 16 CT Scanner (Siemens AG, Erlangen, Germany) was used for all the scanning. The phantom was positioned supine on the couch with a foam headrest for support, as shown in Figure 1a
. Keithley 35050A electrometers (Keithley Instruments Inc., Cleveland, OH) were connected to the ionization chambers to allow simultaneous read-out of fetal and breast dose.
The scan length was set to cover the whole lung field (302 mm) and, each time the patient was re-registered or the phantom was repositioned, a scan projection radiograph (topogram) was used to verify positioning.
All lead aprons used as shielding were checked for any tears, cracks, other damage or movement of the lead before use and were positioned in a straight line on the phantom at the boundary of slices 24 and 25, which marked the position of the lower costal margin. Scan projection radiographs showing the phantom with and without lead can be seen in Figure 2.
|
The lead was positioned with one apron beneath the phantom and one apron on top. The two bottom edges of the apron were lined up with each other and with the phantom.
The following experiments were designed from the results of the protocol audit to determine the effect on fetal dose of scan parameters and the use of lead.
Scan parameters
Standard
An experiment was performed with five repeats using standard parameters: 100 kVp, 140 effective mAs, 16 x 0.75 mm collimation giving 1 mm images, 0.5 s rotation, 10 mm feed per rotation (giving a pitch of 0.833) – this is the protocol that is used, as standard, on this CT scanner. All values of mAs used in this study are effective mAs values as displayed on the Siemens scanner. Effective mAs is defined as the tube mAs per rotation divided by helical pitch. Dose optimisation software was switched off. This experiment was designed to determine the reproducibility of the dose values obtained in the experiments.
kVp
Measurements were made at beam energies of 80 kVp 100 kVp, 120 kVp and 140 kVp.
mAs
The current time product was varied by progressively doubling the mAs setting from 25 mAs to 400 mAs with a constant rotation time of 0.5 s.
Rotation time
Rotation time was set to be 0.42 s, 0.5 s and 0.75 s as these were the steps available on the scanner.
Helical pitch
The doses were measured for helical pitches of 0.5, 1.0 and 1.5.
Collimation
The effect of doubling the beam collimation from 16 x 0.75 mm to 16 x 1.5 mm was observed.
Lead positioning and thickness
Coverage of patient
The effect of varying the amount of coverage of the phantom was investigated by performing one experiment with only the anterior of the phantom covered with the lead shielding.
Altering the volume covered by the lead
A series of measurements were made to investigate the effect of the position of the lead shielding relative to the scan volume. An initial measurement of fetal dose was made with no lead shielding on the phantom. Lead coats were then positioned above and below the phantom such that the superior edges of the shielding aligned with the superior edge of the uterus. The shielding was then moved towards the scan volume (cranially) in 2.5 cm steps, and a measurement of fetal dose was made at each position of the shielding.
Moving scan volume
The effect of moving the scan volume was measured in two separate experiments. First, the whole volume was moved closer to the uterus and, second, the end point of the scan was moved towards the uterus while the top edge stayed fixed. Both these sets of measurements were made in 2 cm steps with and without lead.
Lead thickness
The thickness of lead covering the phantom was varied from 0.25 to 1.0 mm lead equivalent in thicknesses that were readily achievable using lead aprons, and the doses were recorded as a function of thickness. This was performed at all four values of kVp. The attenuation properties of the lead aprons were not assessed as part of this study.
The ion chamber that was used had a minimum detectable dose rate of 0.53 µGy s–1 (Keithley Instruments Inc., Cleveland, OH; manufacturer's manual). For experiments in which a dose rate approaching this limit was expected, a correction factor was applied. This was determined from the experimental data obtained when changing only mAs where a linear response was expected.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
. The responses showed a large range in the exposure parameters that are used for CTPA examination; for example, kVp values ranged from 100–140 kVp, and mAs values varied from 90–210 mAs. Where possible, we used a range of values for each scan parameter that was greater than the range of values that was identified from the questionnaires. One hospital did not scan pregnant patients, and one hospital did not perform PE scans due to the fact that their available scanner was too slow for this application.
|
kVp
For the measurements made with increasing primary beam energy, it was found that the higher the value of kVp, the higher the dose to both the fetus and the breast. The fetal dose was reduced by between 33% and 47% (Figure 3) through the use of a 0.7 mm lead apron for all energies; however, the magnitude of the dose reduction was energy dependent. A constant thickness of lead gave a smaller reduction at higher energies.
|
Rotation time
As rotation time was varied, no significant change in the fetal or breast dose was recorded, although the standard deviation of the results was larger than that found in other measurements.
Helical pitch
Fetal dose increased linearly with increasing helical pitch. When lead was added to the phantom, a reduction of between 41% and 45% was seen at all values of pitch (Figure 4). As the pitch increased, the percentage reduction in dose decreased.
|
Lead positioning and thickness
Coverage of patient
Covering only the anterior of the phantom gave 50% of the dose reduction that was found by covering all around the phantom with a given thickness of lead (i.e. a percentage dose reduction of 
20% with only the anterior of the phantom covered compared with 40% with lead all around the phantom).
Altering the volume covered by lead
The variation in the dose to the fetus as lead aprons were moved progressively closer to the scan volume is shown in Figure 5, which shows that, when the shielding is moved towards the scan volume, a greater reduction in dose was seen. The fetal dose decreased linearly from 56.9 µGy±1% with the superior edge of the shielding placed at the superior edge of the uterus to 25.9 µGy±1% where the superior edge of the shielding was at the inferior edge of the scan volume (a linear fit to the data is shown). The dose (in µGy) at any position of the shielding is given by 56.9–1.4 x distance to uterus (cm) where 56.9 represents the dose when the superior edge of the shielding is aligned with the superior edge of the uterus (the measured dose was 63.3 µGy when no shielding was present).
|
. With the addition of lead shielding, a greater reduction in fetal dose was seen as the scan moved further from the uterus. The edge of the lead apron was placed 150 mm from the uterus, so Figure 6 shows a region in which the lead was irradiated by the primary beam. Moving the whole scan volume closer to the position of the uterus gave an identical increase in fetal dose to moving only the bottom edge of the scan volume. However, the breast dose was reduced from 6.67 mGy±1% to 5.22 mGy±1% over the change in scan position.
|
. For the range of thicknesses shown, the uterus dose reduced as the thickness of lead covering increased.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Experimental results
As expected, the fetal dose increased with beam energy and was reduced with the addition of lead at all energies. Using a lower value of kVp would increase the contrast of the diagnostic images, but also increase patient dose.
The dose to the fetus increased linearly with tube current. With the addition of lead, the fetal dose (µGy) was reduced from 0.5xmAs to 0.3xmAs at 100 kVp. Although using a lower value of tube current would decrease dose, it would also increase the noise in the images.
At low mAs settings, the measured dose deviated from the theoretically linear relationship between dose and mAs despite quality assurance tests on the scanner showing that the measured CT dose index values were linearly related to the mAs setting. This discrepancy was attributed to the measured dose rate approaching the minimum detectable dose rate of the chamber. As such, a correction factor was applied to the dose values at low mAs settings so that corrected dose values described a linear relationship with mAs settings. For all other experiments in which a dose rate approaching this limit was expected, the correction factor was applied to the measured dose values.
No significant changes were seen when the rotation time of the X-ray tube was varied as the value of mAs was kept constant. The small changes that were seen were attributed to mechanical tolerances.
The fetal dose was observed to increase as the helical pitch increased. The scanner that was used compensated for the pitch by changing the current to give a constant level of noise and hence a constant dose level (constant effective mAs). For the reconstruction of helical data, the scan was automatically lengthened as pitch was increased. This change in the distance of the nearest point of internal scattering to the uterus increased the dose by 15.5 µGy±1%. As such, the increase in the fetal dose with pitch setting for constant effective mAs setting is as expected for this type of scanner; this effect may not be seen on scanners that do not automatically adapt the mAs setting when the pitch is changed. Because only the externally scattered radiation is shielded by the addition of the lead apron, as the internal scatter increases, the component that can be reduced by shielding is proportionally less, and so the percentage dose reduction was seen to reduce from 44% to 41%. This indicates the importance of the distance between the end position of the scan and the fetus in determining the dose received by the fetus.
A similar effect was seen when the collimation was increased. Widening the beam brings the closest point of internal scatter closer to the uterus, and this increased the radiation dose. A change in the dose of 1.9 µGy±1% was seen and a change in the percentage dose reduction with the lead apron from 38% to 36%.
When only the anterior of the phantom was covered, the percentage dose reduction was half the value of that when the anterior and posterior of the phantom were covered due to the circular nature of the CT scan. The results indicate that, if a patient is unable to tolerate the weight of the lead placed on the lower abdomen, a significant dose reduction may still be achieved by placing lead underneath the patient and, if possible, using a lesser thickness of lead apron to cover over the abdomen.
Fetal dose varied linearly as the distance between the superior edge of the lead shielding and the superior edge of the uterus increased, as shown in Figure 5. The percentage reduction that was found varied linearly from zero, with no lead present, up to a maximum of 55%. At this point of maximum dose reduction, the lead apron covered the patient up to the edge of the scan volume. The internal scatter from the scan volume is not affected by the lead shielding and is therefore a constant percentage of the total fetal dose. The decrease in the fetal dose that was observed as the lead shielding was moved cranially is a result of the lead shielding blocking the externally scattered radiation that comes from the CT gantry and collimators. As the lead shielding is moved towards the scan volume, a higher percentage of this externally scattered radiation is blocked by the shielding, which therefore yields an increase in fetal dose reduction as the shielding is moved cranially. The results of these measurements show that there is an extra benefit to the patient of covering the greatest volume possible – up to but not covering the scan volume. However, comfort should also be taken into account, particularly if the patient has respiratory difficulties.
The increase in dose that was observed when the bottom of the scan volume or the whole scan volume was moved towards the uterus was exponential, as shown in Figure 6. With the addition of lead, an exponential relationship was still seen in the region where the scan volume did not cover the lead. The percentage reduction in dose is dependent on the distance of the scan volume from the uterus. As this distance increased, the percentage reduction possible increased from 34% to 38%. The implication of these results is that a reduction in fetal dose can be achieved by scanning only as far down the lung field as is necessary for the diagnosis of PE. There will also be an effect of gestational age as the fetus grows towards the scan volume, as shown in studies such as that by Winer-Muram et al [1]. This reinforces the need to minimize scanning of the lower lung field as far as possible. No significant difference was seen in the fetal dose between moving the bottom of the scan volume and moving the whole volume, but a reduction was seen in the dose to the breast due to a reduction in the primary beam irradiating this tissue. This implies that, beyond a certain distance from the point of measurement, the contribution to the fetal dose from internally scattered radiation is negligible.
The variation in the dose to the fetus with thickness of lead at a number of values of beam energy is shown in Figure 7. The thicknesses used were conveniently available in the clinical setting by using lead aprons. As the thickness of lead increased, the dose to the uterus reduced. As can be seen from Figure 7, the dose reaches a point beyond which a further reduction is not seen. This is the point at which the component of the uterus dose that is due to the externally scattered radiation and which has an exponential dependence on thickness has been blocked. When the beam energy was increased, the fetal dose due to both internal scatter and externally scattered radiation was increased. The figures also show that the extra percentage dose reduction gained by adding a thickness of lead greater than 0.35 mm is small (7% at 100 kVp) and that beyond 0.7 mm of lead, the additional dose reduction is negligible (1% at 100 kVp). This implies that the optimum thickness of lead is 0.7 mm (i.e. a doubled-over 0.35 mm lead apron), although use of 0.35 mm of lead would still yield a significant reduction in fetal dose. We found no evidence of lead shielding increasing the fetal dose for thicknesses of lead up to 1 mm.
Clinical protocol
The results have indicated that optimal reduction of fetal dose can be achieved through the use of lead shielding and possibly through some modifications to the protocol used for acquiring the data needed to test for PE. An awareness of the effects of the parameters of the scan on fetal dose can be used to inform decisions about the use of individual scanners when considering pregnant patients.
It is possible to reduce the dose to the fetus by:
The findings of this work are contrary to the results of Hidajat et al [16], who report that no dose reduction was achieved through lead shielding of the uterus. Hidajat et al covered a smaller volume of the patient with the lead apron and performed a scan that was much closer to the uterus. Both these factors have been shown to increase the dose to the uterus and reduce the amount of dose reduction possible through the use of the lead shield. The experimental data collected were used to extrapolate to the position of the scan volume and lead used by Hidajat et al, and the resulting values agreed closely. A slightly greater reduction in dose was seen in the calculated values, but these were based on thicker lead and a lower kVp, which account for the difference.
|
Conclusions |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
|
Acknowledgments |
|---|
Received for publication August 14, 2006. Revision received October 27, 2006. Accepted for publication November 6, 2006.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. K. Pahade, D. Litmanovich, I. Pedrosa, J. Romero, A. A. Bankier, and P. M. Boiselle Quality Initiatives@;DELIM@;Continuing Medical Education: Imaging Pregnant Patients with Suspected Pulmonary Embolism: What the Radiologist Needs to Know1 RadioGraphics, May 1, 2009; 29(3): 639 - 654. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. K. Pahade, D. Litmanovich, I. Pedrosa, J. Romero, A. A. Bankier, and P. M. Boiselle Quality Initiatives: Imaging Pregnant Patients with Suspected Pulmonary Embolism: What the Radiologist Needs to Know RadioGraphics, March 30, 2009; (2009) 293085226. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| BJR | DMFR | IMAGING | ALL BIR JOURNALS |
