BJR
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
British Journal of Radiology (2004) 77, 28-38
© 2004 British Institute of Radiology
doi: 10.1259/bjr/93969091
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by McVey, G
Right arrow Articles by Weatherburn, H
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by McVey, G
Right arrow Articles by Weatherburn, H

Full Paper

A study of scatter in diagnostic X-ray rooms

G McVey, DPhil and H Weatherburn, PhD

Medical Physics Department, The Churchill, Oxford Radcliffe Hospital NHS Trust, Headington, Oxford OX3 7LJ, UK


   Abstract
Top
 Abstract
Introduction
 Methods and materials
 Results and discussion
 Conclusions
 References
 
An important part of determining the radiation protection requirements during X-ray room design is the calculation of the amount of scatter inside and outside the planned locations of the shielding barriers. In this work, a Monte Carlo code has been developed to calculate the percentage scatter so that the current data can be consolidated and new data can be provided as required. Calculations have been compared with measurements to show that they are representative of scatter found in X-ray rooms. Scatter from the dose–area product meter and the collimator system were found to provide large contributions to the measured scatter. A fluoroscopy room containing a C-arm X-ray set was modelled with the Monte Carlo code. The scatter dose was calculated at an X-ray room entrance behind the protective screen, acting as a secondary barrier, at the radiographer's console. The variation of scatter with the position of the protective screen was studied. An empirical calculation, which provided reasonable agreement with the Monte Carlo calculations, was found by using new data for the variation of scatter from concrete with field size. If a door was placed in the X-ray room entrance behind the protective screen, the results showed that it would not need to be lead lined.


   Introduction
Top
 Abstract
 Introduction
Methods and materials
 Results and discussion
 Conclusions
 References
 
The doses at points outside the planned locations for shielding barriers, such as walls and doors, have to be calculated as part of the radiation protection design process for an X-ray room. This design has to take into account primary, scatter and leakage radiation. Appropriate data are obtained from various reports [1–4]. The older reports [1–3] provide data on the amount of scatter and the transmission of primary radiation which were measured using X-ray units with single phase generators. These data need to be updated as the use of modern X-ray units with medium frequency generators is becoming increasingly common. The differences in the penetrating power of the X-rays produced by the different generators prompted Archer et al [5] and Simpkin [6, 7] to produce new transmission data for primary and leakage radiation. These new transmission data are mutually consistent. A recent BIR report [4] presents new scatter data which were measured by Williams [8]. However, this work does not appear to be consistent with the earlier reports. There is considerable variation of Williams' scatter values in the forward and backward directions compared with the older measured values [9, 10]. Therefore, scatter distributions in X-ray rooms are investigated in this paper.

A Monte Carlo code XYZSCAT, based on the EGS4 code system [11], has been developed for this purpose. The scatter from blocks of solid water and concrete have been calculated and compared with measurements carried out in a fluoroscopy room at the Churchill Hospital, Oxford to show that they are representative of scatter in a clinical X-ray room. The comparison was carried out for tube voltages between 49 kV and 121 kV, for scattering angles between 45° and 150° and for field areas between 100 cm2 and 900 cm2. The XYZSCAT code calculates the scatter contributions from different components modelled in the X-ray room, for example, the dose–area product (DAP) meter. This work highlights the significant contributions to the total scatter from materials other than the patient.

Sutton and Williams [4] provide worked examples for a wide variety of shielding problems which may be encountered. However, they only give limited advice on the common shielding problem of how to estimate the dose behind the protective screen at the radiographer's console produced by scatter from the ceiling. Instantaneous dose is difficult to measure at this position. Thus, a practical benefit of Monte Carlo calculations is that they can provide data for these situations. A fluoroscopy room at the Royal Marsden Hospital, London was modelled with the XYZSCAT code and the dose calculated at a room entrance behind a protective screen at the radiographer's console. In the past, an empirical calculation would have been used to estimate the scatter behind the protective screen. This method estimates the scatter from the concrete ceiling as 5% at 1 m [4] using a 10 000 cm2 field area [1]. The XYZSCAT code has been used to provide new data by calculating the scatter from concrete for varying field area. The methods described in this work are generally applicable and may be used for radiographic rooms as well.


   Methods and materials
Top
 Abstract
 Introduction
 Methods and materials
Results and discussion
 Conclusions
 References
 
Output and scatter measurements
Measurements were carried out in a fluoroscopy room at the Churchill Hospital, Oxford. The X-ray generator used was a Philips Medio 65 CP-H (Philips Medical Systems, Eindhoven, The Netherlands)with a Philips overcouch X-ray tube SRO 33 100 with a 13° target angle, 3.4 mmAl total filtration and a 3% peak voltage ripple. All measurements were performed in radiographic mode. The incident air kerma (without back scatter) and half value layers were measured using a Keithley 15 cm3 ionization chamber (Keithley Instruments, OH) connected to a 37D electrometer (Pitman Instruments Ltd, Surrey) for tube voltages between 49 kV and 121 kV. The chamber had a calibration traceable to PTB (Physikalisch-Technische Bundesanstalt), Germany. A DIGI-X instrument using the Keithley chamber, remote controlled with "oRTIgo" software (RTI Electronics AB, Mölndal, Sweden), was used to measure the X-ray tube filtration and voltage waveform ripple.

Figure 1Go shows the experimental set up used to measure the scatter dose. A solid water block (WT1 material) or an aerated concrete block was positioned against one of walls in the X-ray room, at a distance of 100 cm from the X-ray tube focus. The WT1 material was employed as a scatterer to simulate the patient as it is commonly used as a tissue substitute. The aerated concrete block was used as its composition is similar to the materials used in the construction of X-ray rooms although it had a lower density (0.48 g cm-3) than standard concrete. Barytes plaster is required to provide adequate shielding if aerated concrete blocks are used in the construction of an X-ray room. The centre of the field was aligned with the centre of the phantom's incident surface area using the light field. The dimensions of the WT1 block were 30.5 cm long, 30.5 cm wide and 10 cm thick and the dimensions of the concrete block were 44.3 cm long, 21.3 cm wide and 10 cm thick. Measurements of scattered air kerma were carried out with a Pitman 35 cm3 ionization chamber connected to the 37D electrometer. This chamber also had a calibration traceable to PTB. The centre of the chamber volume was positioned 100 cm from the centre of the field size on the incident surface of the phantom. Measurements were obtained at a scattering angle of 135° with tube voltages between 49 kV and 121 kV for an incident field area of 400 cm2 and with a tube voltage of 121 kV for field areas of 100 cm2 and 900 cm2.



View larger version (48K):
[in this window]
[in a new window]
  		Figure 1. The experimental set up used to measure and calculate the scatter from a phantom supported against one of the walls with the X-ray tube focus and detector 1 m from the incident surface of the phantom. DAP, dose–area product.

 
Figure 2Go shows the experimental set up used to investigate the scatter contributions from the X-ray tube head. The solid water block was positioned in the centre of the X-ray room. Measurements of scattered air kerma were undertaken at scattering angles of 45°, 87°, 120°, 135° and 150° with a tube voltage of 69 kV for an incident field area of 400 cm2. These measurements were undertaken with and without a mobile shield placed between the X-ray tube head and the detector. The mobile shield was placed as close as possible to the X-ray tube head to ensure that it did not block any scatter from the phantom reaching the detector. Thus for the measurements at the 120°, 135° and 150° scattering angles, the shield is placed away from the phantom and for the measurements at the 45° and 87° scattering angles, the shield is placed closer to the phantom. Transmission measurements were carried out with the mobile shield to ensure that the scatter from the X-ray tube head was reduced to a negligible value.



View larger version (34K):
[in this window]
[in a new window]
  		Figure 2. The experimental set up used to measure and calculate the scatter from a phantom in the centre of the room and to study the effect of scatter from the X-ray tube head. The positions of the detector and mobile shield are shown for the measurements carried out at scattering angles of 45°, 87° and 135°. DAP, dose–area product.

 
The leakage from the X-ray tube head was also measured and the readings of scatter dose were corrected if it was non-negligible. The percentage scatter was found as the ratio of the scattered air kerma to the incident air kerma without backscatter. The overall uncertainty in this measured quantity was ±11%. The large uncertainty was due to the difficulty in reproducing the position of the chamber in a large dose gradient as the scatter varies rapidly with both scattering angle and distance from the phantom.

The XYZSCAT Monte Carlo code
The XYZSCAT code was developed for this study by modifying the XYZDOS code [12]. Figure 3Go shows an example of the model that can be used with the XYZSCAT code where voxels define a slab phantom and the X-ray room walls. The XYZSCAT code transports photons from the X-ray tube focus to the phantom from where the photons are scattered throughout the X-ray room. Scattering will also occur from other materials in the X-ray room. The program calculates the air kerma in all voxels specified as air. The scattered air kerma is found by combining fluence with the scattered photon energy and the mass energy transfer coefficient [13] for the scattered photon energy. The scattered fluence is calculated by using the tracklength of the photons traversing the voxel and dividing by the voxel volume [14, 15]. A voxel size of 10 cm x 10 cm x 10 cm was used to score the fluence. The radial distribution of scatter is found by using voxels with their centre at 1 m from the phantom surface at scattering angles between 30° and 150°. The percentage scatter is found by dividing the scattered air kerma with the incident air kerma without backscatter.



View larger version (65K):
[in this window]
[in a new window]
  		Figure 3. An example of the voxel geometry that can be used to define the patient (WT1 material), the surrounding air and the X-ray room walls (concrete) with the XYZSCAT Monte Carlo code.

 
The XYZSCAT code was used to simulate four models in this paper which are described below. The XYZSCAT calculations were carried out with between 50 x 106 and 400 x 106 photons divided into 10 batches in order to achieve a statistical uncertainty of ±1% for the simulation of scatter distributions surrounding slab phantoms (models 1 to 3). The statistical uncertainty achieved for the simulation of scatter behind the protective screen (model 4) was ±3%. These calculations were carried out with 240 x 106 photons divided into 10 batches. The XYZSCAT code samples a lagged-Fibonacci random number generator [16] to provide the large number of photon histories required for all of these simulations. The photon histories were followed down to an energy of 1 keV. The interactions simulated included photoelectric absorption with the production of K fluoroescent photons for lead, Compton scattering without electron binding corrections and coherent scattering. Electrons were not transported using the EGS4 code. The photon interaction coefficients, in the energy range between 1 keV and 200 keV, were prepared for use with the XYZSCAT code by the PEGS4 program [11].

Models used with the XYZSCAT code
The first model was used to compare the calculated and measured scatter from blocks of WT1 material and aerated concrete as shown in Figure 1Go. The Philips fluoroscopy room was modelled with dimensions of 6 m long, 6 m wide and 3 m high. The simulated blocks of scattering material have the same dimensions as those used in the measurements. Table 1Go shows the compositions and densities of all materials used in the calculations. The composition and density of the WT1 material was obtained from White et al [17]. The walls, ceiling and floor were simulated as 20 cm thick concrete with a standard density of 2.35 g cm-3. The composition of concrete was obtained from Simpkin [6] with either the measured density of the aerated concrete blocks or the standard concrete density used. The DAP meter (Physikalisch-Technische Werkstätten, PTW type B 727085, Freiberg, Germany), used to monitor DAP, was included in the calculation model. The dimensions of the meter are 16.2 cm long, 18.1 cm wide and 1.7 cm thick. It is constructed from layers of Perspex 0.2 cm thick with an air gap between them. The DAP meter was modelled as a solid block of Perspex with a composition given by the International Commission on Radiation Units and Measurements (ICRU) [18], with an average density of 0.319 g cm-3 found by the mass of Perspex divided by the volume of the DAP meter. The air in the X-ray room was included in the simulation. Its composition and density were given by the National Physics Laboratory (NPL) [19]. The calculations were carried out for tube voltages between 49 kV and 121 kV with a field area of 400 cm2 and for field areas between 100 cm2 and 900 cm2 at a tube voltage of 121 kV.


View this table:
[in this window]
[in a new window]
  		Table 1. The compositions and densities of materials used in the Monte Carlo calculations

 
The second model was used to investigate the scatter from the X-ray tube head as shown in Figure 2Go. The calculations were carried out for tube voltages between 49 kV and 121 kV with a field area of 400 cm2.

The third model simulates the scatter from a concrete barrier (with a standard density of 2.35 g cm-3) with varying field area as shown in Figure 4Go. The calculations were carried out for 140 kV X-rays incident on a 20 cm thick concrete barrier with square field areas increased from 4 cm2 to 40 000 cm2. The calculations were undertaken with a plane parallel beam and with a divergent beam with a focus–surface distance (FSD) of 100 cm.



View larger version (22K):
[in this window]
[in a new window]
  		Figure 4. The geometry for the calculation of scatter at a 150° scattering angle for 140 kV X-rays incident on a 20 cm thick concrete barrier.

 
The fourth model simulates a fluoroscopy room containing a C-arm X-ray set at the Royal Marsden Hospital, London as shown in Figure 5Go. The patient's torso was modelled as a slab of average soft tissue (90 cm long, 22 cm wide and 18 cm thick) supported by a carbon fibre couch top (220 cm long, 56 cm wide and 1 cm thick). The calculation model includes the DAP meter (PTW type B), concrete ceiling, floor and walls (all 20 cm thick, 2.35 g cm-3 density), a false plywood ceiling (1.3 cm thick, 0.3 g cm-3 density) and a protective screen (2 mm lead equivalence). The data for average soft tissue were obtained from the ICRU [20] and the data for wood and carbon fibre were obtained from the NPL [19].



View larger version (24K):
[in this window]
[in a new window]
  		Figure 5. The calculation model of the Philips fluoroscopy room at the Royal Marsden Hospital: (a) the overhead view with the X-ray tube in a lateral orientation and (b) the lateral view with the X-ray tube in the overcouch orientation. The patient access to the X-ray room is not shown in the figure. DAP, dose–area product.

 
Undercouch, overcouch and lateral orientations of the X-ray tube were simulated for the barium enema, barium swallow and other examinations including barium meals and nephrostomy. The X-ray set would not normally be used in an overcouch orientation but it has been included for completeness. Table 2Go shows the imaging parameters for the different examinations and tube orientations. The number of examinations per week represents the workload in a busy department [4]. The fluoroscopy time was obtained from the recommended national reference doses for complete examinations [21]. The incident dose rates at the patient entry point, shown in Table 2Go, are typical for a patient thickness of 20 cm [22, 23]. Large field sizes were employed in the calculations to produce a large amount of scatter and hence provide a conservative estimate of the dose at the X-ray room entrance. For a C-arm X-ray set, there is an image intensifier present behind the phantom which has 2 mm lead equivalence in order to absorb primary radiation. The intensifier has not been modelled so that the results of the calculations are also applicable to radiographic rooms.


View this table:
[in this window]
[in a new window]
  		Table 2. Imaging parameters for the examinations carried out with a Philips C-arm system

 
The amount of scatter reaching the X-ray room entrance is dependent on the position and size of the protective screen at the radiographer's console. Thus, the effect on the scatter of varying the distance of the screen from the X-ray room walls and ceiling and the height of the screen has been studied. The scatter dose was calculated in order to decide whether a door at the X-ray entrance situated behind a protective screen is required to be lead lined. The amount of shielding x in the door required to provide a dose less than a dose constraint Dc of 0.3 mSv per year is found by using Equation 1Go and increasing x until the dose behind the door D(x) is less than Dc. Go


Equation 1Go takes into account the leakage and scatter dose at the X-ray room entrance from the barium enema, barium swallow and the other examinations (De, Ds and Do, respectively) and the corresponding transmissions Te(x), Ts(x) and To(x) because the different examinations have different tube voltages. The transmission data for lead and wood (0.55 g cm-3 density) were obtained from Simpkin [7]. The dose behind a typical door of 4.4 cm thick wood [5] was also calculated using Equation 1Go.

Empirical calculation of scatter behind the protective screen
The XYZSCAT calculations of scatter at the X-ray room entrance behind the protective screen were compared with an empirical calculation described by Equation 2Go [24]. Go


where: Apatient is the field area incident on the patient; Aceiling is the nominal irradiated area of the ceiling; dceiling is the distance from the patient to a position on the ceiling above the protective screen; dentrance is the distance from the ceiling to a point at the X-ray room entrance.

Equation 2Go calculates the percentage scatter, S, in two parts. In the first bracket, the amount of scatter from the patient incident on the ceiling above the protective screen is calculated. The scatter factor is 0.0005 at 1.0 m from a patient for a field area of 100 cm2. Table 2Go shows the values Apatient for each examination and tube orientation. dceiling is 2.96 m. In the second bracket, the scatter from the ceiling reaching the X-ray room entrance is calculated for two positions. The first position is at a gonad height of 0.7 m above the floor (dentrance=3.66 m). The second position is at an eye height of 1.7 m above the floor (dentrance=3.05 m). For simplicity, the same scatter factor of 0.0005 at 1 m from the ceiling is used for a field area of 100 cm2 [1]. Aceiling is chosen from measured or calculated data for the variation of scatter with field size and will be discussed later.

X-ray spectra
The X-ray spectra used in the XYZSCAT code was calculated using the Birch and Marshall Model [25]. The spectra for the Philips X-ray tube were calculated using its X-ray tube characteristics mentioned above. A detailed description of the X-ray spectra calculations is given by McVey [26].


   Results and discussion
Top
 Abstract
 Introduction
 Methods and materials
 Results and discussion
Conclusions
 References
 
Scatter from tissue equivalent phantoms
Table 3Go shows the comparison of the measured scatter with the values calculated using the first model (Figure 1Go) for X-rays incident on the solid water phantom. The scatter is compared at a 135° scattering angle for tube voltages between 49 kV and 121 kV and for field areas between 100 cm2 and 900 cm2. Table 4Go compares the measured scatter with the values calculated using the second model (Figure 2Go). The first set of results in Table 4Go compares the measured and calculated scatter for a 400 cm2 field area for different tube voltages and scattering angles. There is reasonable agreement between the measured and calculated scatter due to the large measurement uncertainty. However, the calculated scatter consistently underestimates the measured value. This is probably due to scatter from the X-ray collimators not accounted for in the XYZSCAT model and is discussed in more detail in the next section. A rigorous validation of the XYZSCAT code is described by McVey [26].


View this table:
[in this window]
[in a new window]
  		Table 3. The comparison of the measured and calculated scatter from the WT1 material phantom for field areas between 100 cm2 and 900 cm2 and for tube voltages between 49 kV and 121 kV

 

View this table:
[in this window]
[in a new window]
  		Table 4. The first comparison (columns 3 and 4) is between the measured and calculated scatter with varying tube voltage and scattering angle. The second comparison (columns 5 and 6) is between the scatter measured with the mobile shield in position and the values calculated with the dose–area product (DAP) meter not included in the computation

 
The measured and calculated scatter is briefly compared with the values given in the literature. For scattering angles less than 90°, the measured and calculated scatter (Table 4Go) tends to be larger than that reported by Trout and Kelley [9] and Williams [8]. This is due to the solid water block used in our work being thinner than the phantoms used in the published work. For scattering angles greater than 90°, the measured and calculated scatter (Tables 3 and 4GoGo) is substantially lower than the values of Williams [8] and greater than those of Trout and Kelley [9] as their experimental arrangement did not include a DAP meter. Our values are lower than those of Williams as in his experimental arrangement the X-ray tube head was closer to the phantom which resulted in a greater scatter contribution from the DAP meter.

Both the measured and the calculated scatter have a similar variation with tube voltage and field area. The scatter at the 135° scattering angle increases by 64% for increasing the tube voltage from 49 kV to 121 kV. There is also a linear relationship between the scatter and field area for slab phantoms. This agrees with the work of Bomford and Burlin [10] for scatter in the backward direction. Trout and Kelley [9] did not find a linear relationship, which was possibly due to their use of a human torso shaped phantom. This implies that the variation of scatter from patients is not linear with field area and needs further investigation [26].

The effect of scatter from the surrounding materials in an X-ray room
The second set of results in Table 4Go shows the measured scatter with the mobile shield in place (Figure 2Go) and calculations carried out without the DAP meter included in the simulation. There is reasonable agreement between the measured and calculated percentage scatter due to the large measurement uncertainty. In Table 4Go, the differences between the measured and calculated scatter are decreased by shielding out the scatter from the X-ray tube head.

The scatter from the X-ray collimators could not readily be modelled with the XYZSCAT code. Therefore, the scatter produced by the collimator system was estimated as the difference between the measured scatter without the shield and the scatter calculated with the DAP meter included in the model. From Table 4Go, the scatter produced by the collimator system and its associated uncertainty are estimated to be between 0.026±0.008% to 0.077±0.033%. Therefore, it is a significant proportion of the total measured scatter. The estimated collimator scatter is similar to the value of 0.04% measured by Trout and Kelley [27].

Figure 6Go shows the calculated scatter from the DAP meter for tube voltages of 49 kV and 121 kV. It varies between 0.025% and 0.085% for scattering angles between 30° and 150°. This variation appears to be independent of tube voltage. The scatter from the DAP meter can also be estimated from Trout and Kelley [27] data. They measured the variation of scatter from different thicknesses of Perspex placed in front of the collimator system. Thus for 4 mm thick Perspex simulating the DAP meter, the estimated scatter is 0.09%, which agrees well with the calculated values.



View larger version (18K):
[in this window]
[in a new window]
  		Figure 6. The variation of scatter from the dose–area product meter at the scattering angle positions 1 m from the phantom surface for tube voltages of 49 kV and 121 kV.

 
Scatter from concrete barriers and blocks
Table 5Go compares the scatter from a concrete barrier calculated using the third model (Figure 4Go) with the values given by the HMSO Handbook of Radiological Protection [1] for scattering angles between 120° and 150°. The average of the HMSO scatter factors corresponds to the value of 0.0005 used in Equation 2Go. The table also shows the comparison of the measured scatter from an aerated concrete block with the values calculated using the first model (Figure 1Go). All values are for a field area of 100 cm2. The HMSO measured values are considerably larger than the other values as they will include scatter from the surroundings and they apply to a range of tube voltages between 100 kV and 300 kV. The calculated and measured scatter from the aerated concrete block is larger than the calculated values from the concrete barrier as they include scatter from the DAP meter. There is very good agreement between the measured and calculated values for scatter from the aerated concrete block.


View this table:
[in this window]
[in a new window]
  		Table 5. Comparison of the calculated and measured scatter at 1 m from concrete barriers or blocks for a 100 cm2 field area

 
Variation of scatter from a concrete barrier with field size
Figure 7Go compares the variation of the HMSO [1] percentage scatter and the XYZSCAT calculated values with field area for a plane parallel beam and a divergent beam with a FSD of 100 cm. The HMSO data were measured for tube voltages between 100 kV to 300 kV at a 160° scattering angle. Figure 7Go shows that the shapes of the measured and calculated curves are very different as the field area increases from 4 cm2 to 40 000 cm2. The calculated scatter values are 6.3% and 4.3% for the plane parallel and divergent beams, respectively, for a field area of 40 000 cm2. The calculated scatter has a linear relationship with field size for areas less than 2500 cm2. The HMSO values show a linear relationship with field size for areas below 4000 cm2 and reach a maximum value of 3.6% for a field area of 18 000 cm2. It is difficult to know the reason for the large difference in the shape of the calculated and published curves as the HMSO report gives very little information on the experimental arrangement used to measure the scatter.



View larger version (19K):
[in this window]
[in a new window]
  		Figure 7. The variation of scatter from a concrete barrier calculated with the XYZSCAT code for a parallel beam and a divergent beam (focus–surface distance (FSD)=100 cm) and the values given by the HMSO [1] for 100 kV to 300 kV X-rays.

 
The variation of scatter at a room entrance behind the protective screen at the radiographer's console
For all studies in this section, a barium enema examination (Table 2Go) was simulated with the X-ray tube in a lateral orientation to the patient. The scatter was calculated at eye and gonad height for an individual standing in the centre of the X-ray room entrance. The radiosensitive organs in the human body all lie within this range of heights. Figure 8aGo shows the variation of scatter at the X-ray room entrance with the depth of the radiographer's console, i.e. the distance between the screen and the wall with the room entrance (Figure 5Go). At the minimum depth investigated of 78 cm, the scatter at eye height is 10% lower and the scatter at gonad height is 40% lower than the scatter values for the screen at a depth of 178 cm. At the maximum depth investigated of 268 cm, the scatter at eye height is 14% higher and the scatter at gonad height is 29% higher than the scatter values for the screen at a depth of 178 cm. There are some oscillations in the scatter for the screen at distances of between 229 cm and 245 cm from the wall. This is possibly due to the large gradient in the scatter distribution surrounding the DAP meter (Figure 6Go). At these positions, the screen is close to the DAP meter, which is at a distance of 236 cm from the wall. The decrease in the scatter reaching the room entrance, shown in Figure 8aGo, may result from the scatter from the DAP meter being locally absorbed by the screen. The increase in scatter at the room entrance for increasing depth is probably due to the position of the screen being closer to the patient—the primary source of scatter. For the X-ray tube in the overcouch and undercouch orientations, the calculations show the expected continuous increase in scatter for increasing depth of the screen. The overall variation in scatter is smaller than might have been anticipated considering the large variation in the area behind the protective screen.



View larger version (36K):
[in this window]
[in a new window]
  		Figure 8. The variation of scatter at the X-ray room entrance behind a protective screen for (a) increasing the depth of the radiographer's console area, i.e. the distance between the screen and the wall containing the room entrance; for (b) increasing the width of the radiographer's console area; for (c) increasing the distance between the protective screen and the ceiling; and for (d) increasing the height of the protective screen.

 
Figure 8bGo shows the variation of scatter at the X-ray room entrance with the width of the radiographer's console, i.e. the distance of the screen from the right hand wall. At the minimum width investigated of 100 cm, the scatter at eye height is 14.3 times higher and the scatter at gonad height is 22.6 times higher than the scatter values for the screen with a width of 270 cm. At the maximum width investigated of 350 cm, the scatter at eye height is 0.92 times lower and the scatter at gonad height is 0.95 times lower than the scatter values for the screen with a width of 270 cm. Thus, screen widths greater than 200 cm provide adequate protection of the X-ray room entrance and as the width decreases then it will no longer be adequate. There is considerably greater variation in scatter with changing the width of the protective screen than the depth of the radiographer's console due to the X-ray entrance becoming increasingly exposed to patient scatter as the width is reduced.

Figure 8cGo shows the variation of scatter at the X-ray room entrance as the height of the X-ray room is varied with the height of the protective screen being kept constant at 200 cm, i.e. the gap between the screen and ceiling is varied. At the minimum gap distance of 53 cm with the suspended ceiling touching the screen, the scatter at eye height is 51% lower and the scatter at gonad height is 42% lower than the scatter values for a gap distance of 141 cm. The scatter at eye height reaches a maximum for a gap of 100 cm which is 4% higher than the scatter value for a gap of 141 cm. The scatter at gonad height is similar for gaps between 90 cm and 220 cm. At the maximum gap distance investigated of 300 cm, the scatter at eye height is 29% lower and the scatter at gonad height is 18% lower than the scatter values for a gap of 141 cm. For gaps between 53 cm and 100 cm, the scatter at eye height increases as the area on the ceiling from which the scatter can reach this position increases more rapidly than the square of the distance from the ceiling. For gaps between 100 cm and 300 cm, the scatter decreases as the area on the ceiling increases less rapidly than the square of the distance from the ceiling.

Figure 8dGo shows the variation of scatter at the X-ray room entrance when the height of the protective screen is increased from a minimum height of 200 cm [4] until it touches the suspended ceiling at a height of 288 cm. With the protective screen touching the suspended ceiling, the scatter at eye height is 74% lower and the scatter at gonad height 61% lower than the scatter values for a screen height of 200 cm. The scatter values at eye and gonad height tend to converge for increasing screen height. This implies that the solid angle into which the scatter from the ceiling can reach both the eyes and the gonads is similar for a screen height of 288 cm.

Scatter dose at a room entrance behind the protective screen at the radiographer's console
Table 6Go shows the total scatter dose in mGy per year at the X-ray room entrance calculated by the XYZSCAT code. The dose for each examination is calculated by the percentage scatter times the incident dose given in Table 2Go and then summed to give the total for each projection. The XYZSCAT calculated dose is compared with the scattered dose calculated empirically by Equation 2Go. The problem with using Equation 2Go is that the area on the concrete ceiling Aceiling has to be chosen. If the whole area of the ceiling was used then the percentage scatter reaching the entrance may be unrealistically large. Therefore in the past, a field area of 10 000 cm2 [1] was chosen which corresponds to a scatter value of 5% at 1 m [4]. This field area was chosen as the scatter values given by the HMSO [1] did not increase for areas above 10 000 cm2 (Figure 7Go). However, the XYZSCAT calculated values increase for field areas above 10 000 cm2 (Figure 7Go) and therefore, a field area of 35 000 cm2 was also chosen for use in Equation 2Go.


View this table:
[in this window]
[in a new window]
  		Table 6. The total scatter doses calculated by the XYZSCAT code and the empirical calculations using field areas on the concrete ceiling of 10 000 cm2 and 35 000 cm2. The annual doses have been calculated for lateral, undercouch and overcouch orientations of the X-ray tube using the total incident dose for each orientation

 
Table 6Go shows the Monte Carlo calculated values are larger than both the empirically calculated values in most cases except for the undercouch orientation for the 35 000 cm2 field area. Empirical calculations have previously been the only method of estimating the dose at the X-ray room entrance and it is of concern that the older data (10 000 cm2 field area) give dose values which are significantly smaller than the Monte Carlo values. The empirical calculation using the larger field area of 35 000 cm2 provides an estimate of the dose at the entrance similar to that calculated by the Monte Carlo method. However, this comparison is only valid for the specific situation shown in Figure 5Go as for example, increasing the screen height would significantly reduce the scatter (Figure 8dGo).

Table 6Go shows that the Monte Carlo calculated doses are larger at eye height than at gonad height which implies that the X-rays are being scattered from the ceiling. The doses are larger for the overcouch orientation of the X-ray tube than for other projections as most X-rays are backscattered from the patient to the ceiling. The scatter from the undercouch orientation is similar to the scatter from the lateral orientation. There was no rapid increase in dose for the undercouch orientation where the primary beam is pointed at the ceiling. This shows that the primary beam is attenuated greatly by the patient. The annual dose due to leakage radiation at the X-ray room entrance behind the protective screen was conservatively estimated to be 0.18 mGy at gonad height. This was empirically calculated using the maximum leakage dose rate of 1 mGy h-1 at 1 m [28] from the X-ray tube assembly and the field area on the ceiling of 35 000 cm2. It can be seen that the dose due to scatter is slightly larger than that due to leakage radiation.

Table 7Go shows the thicknesses of lead and wood that would be required to reduce the scatter and leakage doses to the dose constraint of 0.3 mSv per year. The table also shows the annual transmitted doses through a typical 44 mm thick wood door [5]. A significant amount of shielding is required at eye height and when the X-ray tube is in the overcouch position. However, this is only a theoretical study and the door would not be shielded for these situations. It can be seen that the average doses behind a typical wooden door are close to the dose constraint. Therefore, it would be possible to have a door without lead lining at the room entrance providing that the protective screen is used as a secondary barrier.


View this table:
[in this window]
[in a new window]
  		Table 7. The thicknesses of lead and wood required to reduce the total of the scatter (Table 6Go) and leakage doses at the X-ray room entrance to less than a dose constraint of 0.3 mSv per year for the lateral, undercouch and overcouch orientations of the X-ray tube. The table also shows the total dose behind a typical door constructed of 44 mm thick wood placed in the X-ray room entrance and the average dose behind the door for each orientation of the X-ray tube

 

   Conclusions
Top
 Abstract
 Introduction
 Methods and materials
 Results and discussion
 Conclusions
References
 
Data for estimating the amount of scattered radiation are essential when designing a diagnostic X-ray facility. It has often been observed that there are insufficient data to do this properly [4]. Monte Carlo models provide a method for determining appropriate data [26].

The scatter values calculated by a Monte Carlo code have been compared with the scatter measured in a clinical X-ray room. This comparison was more difficult than expected owing to the scatter from surrounding materials in the X-ray room. The DAP meter and collimator system were shown to give large contributions to the measured scatter in the X-ray room. Further experimental work is necessary to find the source of the small remaining systematic differences between the calculated and measured scatter.

Monte Carlo models are an accurate method for calculating the dose where instantaneous dose measurements are difficult to undertake. The Monte Carlo model has been employed to calculate the dose at the X-ray room entrance behind a protective screen. The current method of estimating this dose is to use empirical calculations. However, this method using HMSO data [1] has been shown to underestimate significantly the Monte Carlo calculated dose. A reasonable empirical calculation may be used by increasing the field area to give results consistent with the Monte Carlo calculated data for the specific situation simulated. However, this may not be the case for other situations, for example a different screen height, as the calculated dose decreases with increasing screen height. Finally, Monte Carlo calculations of dose at the X-ray room entrance show that if a door was placed in this position then it would not need to be lead lined.


   Footnotes
 
Current address for G McVey, North Wales Medical Physics, Glan Clwyd Hospital, Bodelwyddan, Denbighshire LL18 5UJ, UK. Back

Current address for H Weatherburn, Princess Noorah Oncology Center, King Abdulaziz Medical City, PO Box 9515, Jeddah 21423, Kingdom of Saudi Arabia. Back

This work has been supported by a grant from Anglia and Oxford Health Authority. Back

Received for publication November 27, 2002. Revision received June 4, 2003. Accepted for publication August 7, 2003.


   References
Top
 Abstract
 Introduction
 Methods and materials
 Results and discussion
 Conclusions
 References
 

  1. HMSO. Handbook of radiological protection, part 1: data. London: HMSO, 1971.
  2. National Council on Radiation Protection, Measurements. Structural shielding design and evaluation for medical use of X-rays and gamma rays of energies up to 10 MeV. NCRP Report 49. Bethesda, MD: NRCP, 1976.
  3. International Commission on Radiological Protection. Protection against ionising radiation from external sources used in medicine. ICRP Report 33. Oxford: Pergamon Press, 1982.
  4. Sutton DG, Williams JR. Radiation shielding for diagnostic X-rays. Joint report of the BIR/IPEM working party: London: BIR, 2000.
  5. Archer BR, Fewell TR, Conway BJ, Quinn PW. Attenuation properties of diagnostic X-ray shielding materials. Med Phys 1994;21:1499–507.[CrossRef][Medline]
  6. Simpkin DJ. Shielding requirements for constant potential diagnostic X-ray beams determined by a Monte Carlo calculation. Health Phys 1989;56:151–64.[Medline]
  7. Simpkin DJ. Transmission data for shielding diagnostic X-ray facilities. Health Phys 1995;68:704–9.[Medline]
  8. Williams JR. Scatter dose estimation based on dose area product and the specification of radiation barriers. Br J Radiol 1996;69:1032–7.[Abstract/Free Full Text]
  9. Trout ED, Kelley JP. Scattered radiation from a tissue equivalent phantom for X-rays from 50 to 300 kVp. Radiology 1972;104:161–9.[Medline]
  10. Bomford CK, Burlin TE. The angular distribution of radiation scattered from a phantom exposed to 100–300 kVp X-rays. Br J Radiol 1963;36:426–39.
  11. Nelson WR, Hirayama H, Rogers DWO. The EGS4 code system. SLAC Report No. 265. Stanford University, CA, 1985.
  12. Bielajew AF, Rogers DWO. A standard timing benchmark for EGS4 Monte Carlo calculations. Med Phys 1992;19:303–4.[CrossRef][Medline]
  13. Hubbell JH. Photon mass attenuation and energy absorption coefficients from 1 keV to 20 MeV. Int J Appl Radiat Isot 1982;33:1269–90.[CrossRef]
  14. Chilton AB. A note on the fluence concept. Health Phys 1978;34:715.[Medline]
  15. Chilton AB. Further comments on an alternative definition of fluence. Health Phys 1979;36:637.[Medline]
  16. Marsaglia G, Zaman A. A new class of random number generators. Ann Probability 1991;1:462–80.[CrossRef]
  17. White DR, Martin RJ, Darlison R. Epoxy resin-based tissue substitutes. Br J Radiol 1977;50:814.[Abstract/Free Full Text]
  18. International Commission on Radiation Units and Measurements. Tissue substitutes in radiation dosimetry and measurement. ICRU Report 44. Bethesda, MD: ICRU, 1989.
  19. National Physical Laboratory. Radiation transport calculations using the EGS4 Monte Carlo system. NPL Course Notes. Middlesex: NPL, 1993.
  20. International Commission on Radiation Units and Measurements. Photon, electron, proton and neutron interaction data for body tissues. ICRU Report 46. Bethesda, MD: ICRU, 1992.
  21. Hart D, Hillier MC, Wall BF. Doses to patients from medical X-ray examinations in the UK – 2000 Review. NRPB-W14. Didcot, Oxon: NRPB, 2002.
  22. Martin CJ, Sutton DG, Workman A, Shaw AJ, Temperton D. Protocol for measurement of patient entrance surface dose rates for fluoroscopic X-ray equipment. Br J Radiol 1998;71:1283–7.[Abstract]
  23. Gelejins J, Broerse JJ, Hummel WA, Schalij MJ. Reference dose rates for fluoroscopy guided interventions. Rad Prot Dosim 1998;80:135–8.
  24. British Standards Institution. Recommendations for data on shielding from ionizing radiation: part 2, shielding from X-radiation. British Standard 4094, Part 2. London: BSI, 1971.
  25. Birch R, Marshall M. Computation of bremsstrahlung X-ray spectra and comparison with spectra measured with a Ge(Li) detector. Phys Med Biol 1979;24:505–17.[CrossRef][Medline]
  26. McVey GH. Monte Carlo computing applied to X-ray room design. D.Phil. Thesis, University of Oxford, 2002.
  27. Trout ED, Kelley JP. Scattered radiation in a phantom from diagnostic quality radiation. Radiology 1965;85:546–54.[Medline]
  28. National Radiological Protection Board. Guidance notes for the protection of persons against ionising radiations arising from medical and dental use. London: HMSO, 1988.



This article has been cited by other articles:


Home page
 Br. J. Radiol.Home page
G McVey
The effect of phantom type, beam quality, field size and field position on X-ray scattering simulated using Monte Carlo techniques
Br. J. Radiol., February 1, 2006; 79(938): 130 - 141.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by McVey, G
Right arrow Articles by Weatherburn, H
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by McVey, G
Right arrow Articles by Weatherburn, H

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
BJR DMFR IMAGING  ALL BIR JOURNALS