| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Full Paper |
1 Newcastle Mater Misericordiae Hospital, Department of Radiation Oncology, Waratah NSW 2298, 2 University of Newcastle, School of Mathematical and Physical Sciences (Physics) Callaghan NSW 2308 and 3 University of Newcastle, School of Health Sciences (Medical Radiation Science), Callaghan NSW 2308, Australia
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResults and discussionConclusionReferences |
|---|
Differences in entrance surface dose (ESD) and dose at depth (3 cm or 5 cm) were compared for two, overlapping fields centred individually on the chest and abdomen (Technique 1) and one large chest-abdomen field (Technique 2 or babygram). The large phantom was irradiated at 54 kVp, 60 kVp and 70 kVp without additional filtration and at 66 kVp and 70 kVp with a rare-earth hafnium filter. Focus–film distance (FFD) and mAs were adjusted to maintain optical density (OD) on each radiograph.
The baseline dose at 54 kVp and 100 cm FFD was (46±2) µGy. Increasing the tube potential from 54 kVp to 70 kVp without additional filtration reduced the ESD by 27%. However, the addition of a 0.05 mm hafnium filter at 66 kVp further reduced the radiation dose by 13%, to produce an ESD of (28±2) µGy. All contrast details were observable at 66 kVp with hafnium filtration.
Technique 1 may lead to an increase in effective dose due to field overlap, which diverges at depth, and increased scatter at the periphery of the fields.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionReferences |
|---|
The National Radiological Protection Board (NRPB, UK) defines entrance surface dose (ESD) as the absorbed dose to air at the point of intersection of the X-ray beam axis with the entrance surface of the patient, including backscattered radiation [2]. The CEC also provides a reference ESD of 80 µGy for an AP chest examination on a neonate of 1000 g. In addition to the CEC reference dose there have been a number of studies that have assessed neonatal dose, and the values are seen to fall between 44 µGy and 92 µGy [3–6]. These doses were assessed at tube potentials ranging from 46 kVp to 61 kVp, with the lower ESDs generally being observed at the higher tube potentials.
In order to reduce the effective dose from any X-ray examination, a number of variables can be manipulated; two prominent factors are a reduction in ESD and increasing collimation or decreasing the volume of tissue irradiated. In neonatal radiography, the radiographer should collimate precisely to the region of interest, being careful not to exclude relevant anatomy.
The patient ESD is an important factor affecting the radiation risk, and can provide a realistic, easily measured guide to the relative level of patient protection being provided by the imaging techniques. The factors affecting ESD are, however, not as straightforward as the use of collimation. Two of the ways to decrease the ESD while maintaining adequate penetration (and hence optical density (OD) on the radiograph) are either to increase the tube potential (kVp) or thickness of beam filtration, or both. Erbium and hafnium are the two most common filter types used in neonatal radiography, with the use of erbium being reported to reduce radiation dose by as much as 50–60% [7–9]. No paper was found that quantifies the dose reduction with hafnium filtration, however a reduction was documented by Adams [9].
Radiation detectors commonly used in medicine are limited in their ability to perform accurate dose measurements in neonatal radiography due to their lower measurement limit, which is typically greater than 50 µGy, and an over-response to low energy X-rays. LiF:Mg,Ti or Li2B4O7 thermoluminescence dosimeters (TLDs) have a lower measurement limit of 50 µGy to 100 µGy, whereas high sensitive materials such as Al2O3:C and CaSO4 TLDs over-respond to low energy X-rays by 3 to 12 times, respectively, at 30 keV relative to MeV energies [10]. Hence, most studies investigating ESD in the neonatal intensive care unit have involved indirect measurements, i.e. dose area product meters operating at ±20% uncertainty, [11, 12] or calculations using radiographic parameters [6–8]. Alternatively LiF:Mg,Cu,P TLDs, only recently applied to neonatal radiography [13, 14], offer increased sensitivity and a more uniform energy response compared with other dosimeters. The lower measurement limit is 1 µGy and the relative energy response is approximately unity except for an under-response of 10% to 20% at X-ray energies less than 20 keV and between 80 keV and 300 keV [15, 16].
Other similar studies have only performed centre-of-field measurements for each imaged field and have not investigated the magnitude of increased dose to radiosensitive organs at the point of overlap (if any) or at the field periphery [6, 12, 17].
The aim of this study was to use an ultrasensitive TLD material to investigate the effect of different collimation techniques (separate versus combined chest and abdomen fields) and effective energy (kVp and filtration) on the ESD to the neonate.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionReferences |
|---|
X-ray unit
All exposures were performed with the mobile X-ray unit (GE AMX4) (General Electric Medical Systems, Milwaukee, USA) used in the hospital for neonatal chest radiography. A full quality control test was initially performed on the unit to assess for linearity, filtration and output consistency, and all results were within acceptable limits [18].
TLDs
Radiation dose was measured using 4.5 mm diameter x 0.8 mm thick LiF:Mg,Cu,P circular detectors (GR-200A, Solid Dosimetric Detector and Method Laboratory, Beijing, China). Annealing was performed in a microprocessor-controlled annealing oven (Scientific Equipment Manufacturers, Magill, Australia) at 240°C for 10 min followed by a quench on an aluminium block partially immersed in water. TLDs were read out in a Harshaw Bicron 5500 automatic TLD reader (8°C s-1, 240°C, 40 s). The TLDs were calibrated at two dose levels, 0.5 Gy and 1 Gy, at 60 kVp (effective energy 35 keV, half-value layer (HVL) 1.6 mm Al) using a superficial therapy unit with a calibration traceable to Australian Radiation Laboratories (National Standard Laboratory). Low doses were inferred from a linear fit extrapolated to zero. Linearity at low dose levels was verified in previous experiments [13]. Accuracy of the output of the superficial therapy unit was within ±3%.
Neonate phantoms
Two neonate-mimicking phantoms were manufactured from tissue-equivalent material (white water, PTW Freidburg, Germany). The inside of both phantoms can accommodate soft tissue-equivalent or lung-equivalent inserts (PTW Freidburg, Germany) to alter torso tissue density. One lung insert was inserted into the small phantom and two for the large phantom. Phantoms were designed to represent neonates of weights 700 g and 2000 g. Figure 1
shows a schematic drawing of the dimensions of the small phantom. Similarly, dimensions of the large phantom include 100 x 250 x 70 mm3 (width x length x thickness), with insert dimensions of 60 x 80 x 20 mm3.
|
Image quality tools can be inserted into a compartment in both phantoms (hatched area in Figure 1
). The contrast resolution tool was manufactured from tissue-equivalent white water, and contains holes of different depths, with diameters varying from 1 mm to 8 mm. In addition, a line pair gauge can be inserted into the large phantom compartment, but spatial resolution was not investigated in the present study.
To authenticate the phantoms as realistic replacements for neonates, the OD of the lung-equivalent insert in the large phantom radiograph produced at typical exposure factors was compared with that of a typical neonatal radiograph.
Rare-earth filter
The rare-earth filter chosen for this study was 0.05 mm of hafnium. Hafnium has a L-edge of 11.3 keV, which reduces lower energy X-rays, and a K-edge of 65.3 keV [19].
Dose assessment for changes in collimation
The TLDs were exposed in the phantom at 100 cm focus-film distance (FFD). For each phantom, TLDs were irradiated at 54 kVp (effective energy 37 keV, HVL 2.6 mmAl) for both field combinations. Field configurations for imaging the chest and abdomen (labelled Techniques 1 and 2) are illustrated in Figure 2. In Technique 2 the one exposure encompasses the entire chest-abdomen region and in Technique 1, two separate chest and abdomen fields generate a region of overlap hence cumulative exposure. For the small phantom, individual field sizes for Technique 1 were both 10 cm x 10 cm at the entrance surface. Total field size was 10 cm x 18 cm, i.e. 2 cm overlap. For the large phantom, individual field sizes for Technique 1 were both 12 cm x 15 cm and the total field size was 12 cm x 20 cm, i.e. 10 cm overlap, at the entrance surface. For Technique 2 the field sizes were 10 cm x 18 cm and 12 cm x 20 cm at the entrance surface for the small and large phantom, respectively. All of these fields were representative of those routinely used in the neonatal unit. Both Techniques 1 and 2 are used in the neonatal unit, with Technique 2 being more common.
|
Dose assessment for changes in beam energy
The exposure factors representing routine clinical conditions for the large neonate were set (54 kVp, 2.5 mAs and 100 cm FFD) and a radiograph was taken of the phantom. The radiograph was developed and the OD of the lung-equivalent insert was found to be comparable with a clinical chest radiograph.
Before any TLD measurements were performed, a series of radiographic exposures were made of the phantom with different exposure parameters (kVp, mAs and FFD) and filtration levels, to assess the correct factors needed to maintain an OD comparable with that of the baseline. The exposure conditions used and resulting ODs are given in Table 1.
|
Assessment of image quality
The modular design of the phantom enabled the investigation of changes in radiographic contrast with changes in beam energy. Contrast was assessed two ways, initially each film was viewed to assess if all of the contrast details were visible, and secondly, the OD was measured in two places on each radiograph produced. One area was the background on the contrast tool, the other was inside one of the contrast disc areas (with 8 mm diameter). Contrast was then simply taken as the percentage difference between the two readings.
|
Results and discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionReferences |
|---|
. For Technique 1, dose variability, i.e. large standard deviation, in measured dose values is due to field inhomogeneity caused by the partially overlapping fields. There is a factor of 1.7 increase in dose at the centre of the field due to field overlap compared with the dose at the field edge. For Technique 2 the dose profile is relatively homogeneous.
|
Changes in beam energy
The results for the large neonatal phantom (2000 g) have been summarized in Table 1. Our baseline dose of 46 µGy is at the lower end of previously reported studies [3–6, 20] and corresponds with the median value for the 1000 g newborn reported by the CEC [1].
Increasing the tube potential from 54 kVp to 70 kVp without any rare earth filtration results in an ESD reduction of 27%. The addition of the hafnium filter at 66 kVp is seen to have a significant effect with the ESD being reduced by a further 13%. However, as the tube potential is increased from 66 kVp to 70 kVp with the hafnium filter in position, the ESD is increased very slightly (6%). This is within acceptable limits considering the greater OD of the film and measurement accuracy of TLDs. Increasing the potential from 54 kVp to 66 kVp and adding hafnium filtration reduced the dose at 5 cm depth by 26%.
Changes in contrast
The measurement of contrast gave the expected results with the reduction of contrast with increasing beam energy as shown in Table 1. There was a rapid reduction from 54 kVp to 60 kVp, but there then appears to be a flattening of contrast as the energy further increased to 66 kVp. Overall there was a 20% reduction in contrast, i.e. percentage difference in OD readings, by increasing the tube potential from 54 kVp to 66 kVp. The introduction of the hafnium filter at 66 kVp saw the contrast further reduced by 7%. While a 20% reduction seems high, visual inspection of the films was carried out under standard clinical conditions and all contrast details were clearly visible on each radiograph.
TLDs
LiF:Mg,Cu,P TLDs had a reproducibility of ±5% (1 SD) down to 20 µGy, which includes the reproducibility of the mobile X-ray unit output. In contrast, dosimeters commonly used in diagnostic radiology are limited in one or all of their properties in providing this dose information, as stated in the introduction. Jones et al used a dose area product (DAP) meter and thus had to expose their phantom 10 times to gain enough signal to derive the effective dose accurately [12]. This of course could not be done in vivo, therefore Lowe et al tolerated ±20% uncertainty in their in vivo DAP readings [11].
Direct, accurate dose measurements during individual radiographs can assist a department in justifying technique modifications in line with CEC recommendations that may provide a substantial dose saving without compromising image quality and hence the diagnosis or monitoring of the neonate's medical condition.
|
Conclusion |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionReferences |
|---|
This study concludes that dose reduction can be easily achieved in our environment. It is recommended that a tube potential of 66 kVp with a 0.05 mm hafnium filter is used for neonatal chest radiography. This tube potential is just outside the upper level (65 kVp) recommended by the CEC, and produces an ESD of 28 µGy, which is considerably lower than the reference level of 80 µGy suggested by the CEC.
When it is necessary to radiograph both the chest and abdomen of a neonate, a significant dose reduction can be achieved by using Technique 2, thus exposing the chest and abdomen in the one exposure, rather than two separate exposures. This dose reduction is a consequence of eliminating field overlap, which diverges at depth, and decreased scatter.
While dose reduction is important, accurate diagnosis relies upon the radiologist being presented with an image of acceptable radiographic quality. The increased beam energy will undoubtedly cause some loss of contrast and it is important to ensure that the image is not significantly degraded. Preliminary results reveal a reduction in contrast as expected, but the actual clinical significance of this reduction is as yet unsure. This is currently being investigated in the clinical environment with radiographs of actual neonates and not phantoms, following CEC criteria.
The change in position of the central ray for Technique 2 will cause a distortion of anatomical structures that will increase with distance from the centring point. Again it is important to be sure that distortion of the size, shape and position of structures does not cause the image quality to be significantly degraded.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received for publication June 17, 2002. Revision received September 9, 2002. Accepted for publication February 6, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S J Mutch and S D P Wentworth Imaging the neonate in the incubator: an investigation of the technical, radiological and nursing issues Br. J. Radiol., November 1, 2007; 80(959): 902 - 910. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E D Dougeni, H B Delis, A A Karatza, C P Kalogeropoulou, S G Skiadopoulos, S P Mantagos, and G S Panayiotakis Dose and image quality optimization in neonatal radiography Br. J. Radiol., October 1, 2007; 80(958): 807 - 815. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| BJR | DMFR | IMAGING | ALL BIR JOURNALS |
) represent beam divergence. Circles along the long axis of the phantom represent recesses on the phantom surface that hold thermoluminescence dosimeters. FFD, focus–film distance.