| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Full Paper |
1 Interfaculty Reactor Institute, Delft University of Technology, Mekelweg 15, 2629 JB Delft and 2 Department of Radiology, Leiden University Medical Center, PO Box 9600, 2300 RC Leiden, The Netherlands
Correspondence: F W Schultz, Interfaculty Reactor Institute (IRI/ST-MSF), Mekelweg 15, 2629 JB Delft, The Netherlands
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResults and discussionConclusionsReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResults and discussionConclusionsReferences |
|---|
X-ray guided cardiac catheterizations are a form of radiology that is applied to examine and treat paediatric patients with congenital heart defects relatively often. High levels of radiation exposure may occur due to using biplane X-ray equipment and prolonged fluoroscopy time and cine imaging to monitor and record access of the flexible wire to the diseased region after entering via a peripheral blood vessel. As children are relatively more radiosensitive than adults [2] and have longer mean lifetime expectancy, it is very important to optimize procedures and keep the dose to the paediatric patient as low as possible. On the other hand, the occupational dose to the medical staff tends to be high in comparison with other radiological procedures. Furthermore, it may build up quickly because the same personnel will be involved regularly in similar catheterizations. The risk to the cardiologist is most probably highest, as this person remains longest with and nearest to the patient during the procedure.
Effective dose, a weighted sum of organ doses [2], is considered the best dosimetric quantity for estimating the risk of exposure to ionizing radiation. To compute effective dose it is necessary to know the organ doses with good accuracy. Direct measurements are very difficult if possible at all. The usual approach is to take a dosimetric quantity for which the value is simpler to establish, e.g. air kerma free in air (Ka), entrance skin dose (ESD) with or without backscatter, or dose–area product (DAP). The organ doses are obtained subsequently by multiplying by an appropriate dose conversion factor (DCF). DCFs depend on the exposure conditions and are generally determined through Monte Carlo (MC) simulation of radiation transport.
As a result of elaborate and careful development of software and the advance of cheaper and faster computers, MC simulation has become a well-accepted technique for solving radiation transport problems in a multitude of applications [3] including medical physics. One of the general-purpose codes is MCNP, which originates from the Los Alamos National Laboratory [4] and has been made widely available. Owing to its great flexibility with respect to the type of application, a relatively large effort from the user is required to construct the proper input and retrieve the desired output information.
In contrast, a commercially available MC code like PCXMC [5] is designed to serve just one type of application, i.e. the calculation of patient doses (effective and organ) in diagnostic radiology. Therefore, the variation in the configuration of input and output variables is limited, which enables a user-friendly interface with an interactive menu structure. Furthermore, by allowing a simplified treatment of the physics, the required processing time for the actual MC simulations can be kept very short.
At the Leiden University Medical Center (LUMC) radiation exposure to patients and staff during paediatric heart catheterizations has been investigated. Patient doses due to various procedures were determined by measuring DAP and using DCFs obtained with PCXMC. PCXMC was chosen because a considerable number of DCF sets had to be evaluated, which would have been too time-consuming with MCNP.
The main results of the investigation are presented elsewhere [6]. The present paper is meant to demonstrate that the choice of PCXMC for calculating the DCFs is justified. To that purpose the more extensive code, MCNP is applied to a small sample of cases. The results are compared with those from PCXMC.
Furthermore, the present paper looks deeper into the radiation burden of the cardiologist, being the staff member who collects the largest dose. In [6], doses to staff are expressed as readings from an electronic personal dosimeter (EPD). For one procedure, DCFs for the scattered radiation are calculated with MCNP, as PCXMC is not suitable for this type of problem. The influence of several shielding measures (apron, collar) is examined. Effective dose is compared with the EPD readings and with calculated values of the personal dose equivalent for penetrating radiation, HP(10) [7]. It is important to know to what extent these quantities agree, because the measured value is assumed to represent the personal dose equivalent, which, as an operational quantity, is in turn supposed to reflect effective dose.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResults and discussionConclusionsReferences |
|---|
For this paper MCNP version 4C is used, which has been implemented on a Compaq XP900 Alpha workstation (Compaq Computer B.V., Gouda, The Netherlands. Unix operating system). Without attempting optimization, i.e. application of additional variance reduction techniques, it typically takes some 6 h of computer processing time (20 million starting particles) to yield less than 0.5% relative statistical uncertainty in the calculated effective dose conversion factor for patients. To calculate the effective DCF for the cardiologist exposed to radiation scattered from the patient, about four times as much processing time is required to yield approximately 4% uncertainty.
PCXMC, developed at the Medical Radiation Laboratory of the Finnish Radiation and Nuclear Safety Authority [5, 8], is a special purpose code for calculating patient dose in diagnostic radiology from either measured dose–area product or entrance air kerma (without backscatter). Operating under Windows (95/98/NT) on a personal computer, the code is fast (minutes) but transports only photons in a limited energy range (up to 150 keV). Focus-to-skin distance (FSD), degrees rotation about the patient's longitudinal axis and in craniocaudal direction, skin entry point of the central beam axis and field size at this position determine shape, position and orientation of the beam. A graphical interface facilitates the beam positioning. A hermaphrodite mathematical phantom represents the patient. In addition to adults, one of five age classes can be chosen for paediatric examinations. To some extent a standard phantom can be adapted to the actual length and weight of the patient. Doses to the eye lenses are not computed. This may be a disadvantage when evaluating examinations with beams in the head region.
Description of the geometry for MCNP
All components of the geometry are placed within an air-filled box. In the practical situation the patient is placed in supine position on the centreline of a table (not actually modelled for the radiation transport simulations) of adjustable height (Figure 1
). The width of the table is 55 cm, including the mattress. The cardiologist leans against the table at the right hand side of the patient, at the level of the groin (or more toward the feet for shorter patients). Thus, the patient's heart can be imaged either from left to right (LLAT) using the lateral tube – to the left of the cardiologist – or from back to front (PA) with the undercouch tube of the biplane X-ray system (Philips Diagnost, Philips Medical Systems, Best, The Netherlands). Imaging is not always performed using exact PA or LLAT projections (as defined in [9]) but deviations of some 10–20 degrees may occur. The undercouch tube can be rotated (if to the right side of the patient this yields a LAO, left anterior oblique view) and/or tilted toward the head (positive angle) or the feet. The lateral tube can also be rotated toward a LAO view. Only the focus of the X-ray tube is modelled, i.e. as a standard MCNP point source emitting a cone of photons (of which the cone axis and top angle can be adjusted). Photon stopping material is used as a collimator model to reduce the circular cross section of the cone to a rectangle. Actual FSD, entrance point of the central beam axis on the patient and field size in the entrance plane determine the pyramid shape of the simulated X-ray beam. In reality a calibrated Diamentor M4 DAP meter (PTW, Freiburg, Germany) mounted on the X-ray machine measures the characteristic physical quantity for normalization in the dose calculations. In the simulations the DAP is calculated as the dose in air averaged over the beam area at, arbitrarily, 20 cm from the focus, multiplied by the beam area at this distance. The mathematical anthropomorphic phantom ADAM [10] represents the cardiologist (male adult). It contains all organs that are required for the calculation of effective dose. One of the paediatric MIRD phantoms [11] is chosen to represent the patient. Based on gender, body weight and age of the patient, a male or female phantom from the age class newborn, 1, 5, 10 or 15 years is selected. Composition and density of the tissues are the same as for ADAM. For lateral views the arms are removed. The phantoms, largely described by first and second order equations, are adapted to the input format of the MCNP code.
|
X-ray spectra
The relative intensities of the photons emitted from the point source, as a function of the energy, must be known for MC simulation. The X-ray spectra to be used in the MCNP simulations are generated with spectrum generator software [12], taking the X-ray voltage and Al filtration applied in the clinic as input parameters. The PCXMC code automatically generates its own spectra, based on the same input parameters.
A spectrum of scattered X-rays was required for the assessment of the dose to the cardiologist. It was obtained with MCNP by recording the energies of photons passing a surface between the patient and the cardiologist, when the former – in this case the newborn male phantom – is being exposed (PA) to the primary beam.
Patients and staff
The procedures to assess and treat congenital heart defects comprised diagnostic catheterizations (n=24), atrial septic defect (ASD) closure (n=14), radio-frequency (RF) ablation (n=10), balloon dilatation (n=11) and patent ductus arteriosus (DA) occlusion (n=10). The mean age of the patients is 5 years and weight 19 kg with ranges of 0–18 years and 2–68 kg, respectively.
Four types of imaging are distinguished for each procedure, i.e. posteroanterior (PA) and lateral (LLAT) continuous fluoroscopy, and PA and LLAT cine exposures for angiography (at 12.5 or 25 frames per second). The DAP during each view was recorded, as were the parameters needed for MC input.
Although four categories of specialists can be distinguished, i.e. cardiologist, echo-cardiographer, anaesthetist and nurse, only the first is considered here as they are assumed to be the most exposed. Various types of protective clothing are worn (lead equivalent apron: short or long, frontal or wrap-around; lead equivalent collar, sometimes). In contrast to the practice in the UK, the radiation exposure is measured with an EPD outside the apron.
|
Results and discussion |
|---|
|
TopAbstractIntroductionMethodsResults and discussionConclusionsReferences |
|---|
. The technical (tube voltage, filtration) and geometric (direction, FSD, field size) parameters of the beam are also presented, for the four views per procedure (LLAT or PA in cine or fluoroscopy mode). Next in Table 1 the DCFs per view for effective dose are shown with their relative (statistical) uncertainty. The results from simulations with the MCNP and PCXMC codes are compared in terms of a ratio (which ideally should tend to 1) and as percent relative difference. MCNP tends to yield slightly lower DCFs. The relative difference has a median value of -2.5% (range 0.6% to -11%), and (in absolute sense) is larger than 10% in only two cases. As DAP per view was actually measured, the DCFs can be used to calculate the effective dose per view and per procedure, according to MC code. These results are also shown in Table 1. For total effective dose per procedure, the relative difference between MCNP and PCXMC is not larger (absolutely) than approximately 5%. From the viewpoint of radiation protection, disagreement at such level is not unacceptable, hence it can be concluded a posteriori that the choice of PCXMC for faster results is justified.
|
, i.e. those for which the codes yield (absolutely) the smallest and largest difference in effective dose. These are, respectively, ASD closure/PA fluoroscopy and RF ablation/LA cine view. In Table 2 the corresponding DCFs for separate organs and for effective dose are listed with their statistical uncertainties. The latter always are larger for PCXMC owing to fewer particles being simulated in this code. Again, for comparing the two codes the ratio and the relative difference are used. Ratios often deviate from the ideal value of 1 (the codes yielding equal results), even for the view for which both codes agree best on the effective dose. Several relative differences in organ DCF between the two MC codes may also seem rather large. This holds both for organs inside (e.g. heart) and outside (e.g. small intestines or kidneys) the primary beam. Except for the lowest doses, (organs far from the beam, like testes and urinary bladder, and therefore the least reliable results), corresponding organ DCFs remain of comparable magnitude. Thus, it can be concluded that both codes yield similar dose distributions over the body. Furthermore, in the present cases only a few organs (lungs, oesophagus, bone marrow) contribute substantially to the effective dose, which further limits the influence of many of the larger differences.
|
Dose to the cardiologist
Using two large-size phantoms in an MC simulation, i.e. one patient and one cardiologist, with all necessary internal organs to be defined, puts a substantial load on the computer resources. By its design purpose, PCXMC does not even allow such a configuration. MCNP does, but its performance would improve if a less complicated set-up of structures in the simulation were allowed. Therefore, tests were made to see if the cardiologist's organ doses (complete phantom) could be calculated with a simplified patient phantom. Only the external shape of the 5-year-old male phantom was used. The body was filled homogeneously with polymethylmethacrylate (PMMA). The PA fluoroscopy exposure of the simplified phantom was compared with the complete patient phantom. No protective clothing was used. The difference in calculated HP(10), just below the centre of the cardiologist's right clavicle, was only 4% (normalized to DAP: 3.07 and 2.96 µSV (Gy.cm2)-1, respectively, for simplified and full patient). However, effective dose to the cardiologist differed by 49% (4.17 and 2.80 µSv (Gy.cm2)-1 for simplified and full patient, respectively). Scatter from the cardiologist himself causes about 22% of the calculated HP(10) value, coming mainly from muscle (9%), skin (5%) and bone, i.e. ribs: 1.5%. The remaining 78% of the dose is due to scatter from the patient. In the detailed phantom the principal contributors are muscle (46.5%), lungs (11%), heart (7%), ribs (4%), liver (4%) and skin (2%). It was concluded that the patient phantom's composition has a strong impact on the scatter properties. Hence, the complete patient phantom had to be used.
The radiation burden to the cardiologist due to two heart catheterization procedures is shown in Table 3. If the cardiologist does not wear protective clothing he receives an effective dose that is either less than 1% (ASD closure) or less than 0.2% (diagnostic procedure) of the effective dose to the patient. The main contributions to the effective dose of the cardiologist come from testes, stomach, colon, bladder and lung doses. The percentages are view-dependent. The calculated HP(10) is only a rough approximation of the effective dose. In the case of ASD closure there is an overestimate (relative difference 19%). For the diagnostic views there is an underestimate (relative difference -43%), which from the viewpoint of radiation protection is worse than overestimation. It should be noted, however, that in this case the internal HP(10) values are listed. As for the other type of procedure, calculated HP(10) just outside the skin overestimates E (relative difference 79%). The reading of the EPD underestimates the effective dose in the absence of protective clothing. This may be caused by non-uniform directional sensitivity and/or insensitivity to lower X-ray energies.
|
As expected, by applying a 0.5 mm Pb equivalent frontal apron (type 1) in the case of diagnostic catheterization the effective dose to the cardiologist is dramatically reduced. The distribution of organ doses also changes, with principal contributions now coming from bone marrow, skeleton, skin, thyroid and brain. Thus, the EPD's above-mentioned underestimation of effective dose to the cardiologist changes to a 100 fold overestimation.
It is obvious that a correction must be applied to the reading of an EPD worn on the apron to obtain a proper indication of the magnitude of the effective dose. Huyskens et al [14] introduced the divider, multiplier and protection factor as characteristic quantities and calculated values for various exposure conditions with regard to shielding and beam orientation. In their model the assumption of a broad unidirectional X-ray beam was made. The divider is the number by which the individual dose equivalent measured on the apron has to be divided to obtain effective dose. The multiplier is a number by which the individual dose equivalent measured under the apron has to be multiplied to obtain the effective dose. Protection factor is the ratio of effective doses without and with the apron being used, which indicates the effectiveness of the shielding.
These quantities are listed in Table 4 for various types of aprons and lead equivalent thickness. The quantities from which they are derived, i.e. HP(10) on and under the apron and effective dose (all normalized to DAP) are also shown. The PA fluoroscopy view of the diagnostic procedure performed with the male baby patient was chosen because it yields the highest contribution to the effective dose of the cardiologist.
|
two beams are considered. Beam A consists of radiation scattered on and in the patient when the latter is exposed to the primary beam from the X-ray unit. This represents the actual situation of exposing the cardiologist. Beam B is rather "artificial" in this respect, being a broad unidirectional beam like the one used by Huyskens et al Beam B is aimed to be incident on the cardiologist at a 45° angle coming from front left (RPO, right posterior oblique view). The photon energies are according to the spectrum of scattered X-rays. The simulations with Beam B were conducted to establish that discrepancies in results observed for the present Beam A and Huyskens' results can be explained by the differences in the beam properties. As can be seen in Table 4 the values of the characteristic quantities for Beam B closely match the ranges stated by Huyskens et al. If no apron is worn, the effective dose calculated with the unidirectional Beam B is lower than for exposure to actual scatter as calculated with Beam A (1.52 versus 2.48 µSv (Gy cm2)-1). When an apron is worn the opposite is true (e.g. 0.21 versus 0.06 µSv (Gy cm2)-1 for the 0.25 mm lead-equivalent wrap-around apron). Thus, considering unidirectional exposure results in underestimating the protection that the apron actually offers. Depending on the apron properties, the protection factor is 2.5 to 7 times higher according to the more realistic simulation than when using Beam B. For instance the lower regions from which the scatter beam originates reduces the contribution of the thyroid to the effective dose. Consequently, the actual value of the quantity divider is higher. In other words, the overestimation of E by HP(10) measured on the apron is increased by a factor of 2 to 5 when exposure is approximated with a unidirectional beam.
The characteristic quantity that applies to the situation in the UK, where staff wear dosimeters underneath protective clothing, is the multiplier. According to the present study the dosimeter reading should be multiplied by 6 and 2.5 to yield appropriate estimates of effective dose for frontal and wrap-around aprons, respectively. The present values differ to some extent from the ones stated by Huyskens et al (Table 4), depending on apron and exposure conditions.
For a number of fluoroscopy procedures (70–110 kV) Faulkner et al [15] have measured doses with TLDs in and on a Rando-Alderson phantom exposed to radiation scattered from another anthropomorphic phantom in the primary beam. For commonly used aprons they found divider values of 2–60 and multiplier values of less than 7. The present study, under different exposure conditions (e.g. 59 kV), is in agreement with their results.
If no apron is used, the on-apron value of HP(10) in Table 4 is lower than the under-apron value. The difference is due to the fact that the under-apron position is on the skin of the phantom and the on-apron position is 2 cm away, i.e. on the spot where the apron would be if it were present. The closer to the phantom, the larger is the backscatter component to the dose. Hence the under-apron value is higher.
That backscatter from the phantom is effectively stopped by the aprons can be deduced from the constant, lower value of HP(10) on the apron with the aprons present compared with the absence of an apron. (Note that the layer of lead itself will not generate much scattered radiation at 60 kV.)
It might be concluded, wrongly, from looking at the HP(10) under-apron values that only thickness and not the type of apron (frontal or wrap-around) matters. However, the influence of the type of apron is evident from the effective dose. Obviously, wrap-around and thicker aprons offer more protection. As expected, addition of a thyroid collar causes a further, though relatively small, improvement.
Padovani et al [16] have evaluated a number of empirical rules that are meant to estimate occupational effective dose from measured operational quantities.
The first holds for a single dosimeter positioned at the neck, measuring HP(10) on the apron. In that case, the divider value of 21 is to be used [17]. The shape of the Adam phantom is not well suited to calculate the individual dose equivalent at the neck in the situation that the radiation source is close to the phantom and has a low frontal position. The protruding trunk then forms an unnatural shield, hiding the "monitor" spot. Assuming that, in reality, the position – on-apron, either high on the chest or on the neck – does not affect the reading of the dosimeter too much, in the present study the rule is about correct for the 0.25 mm Pb equivalent frontal apron. Applying the rule to the other aprons would overestimate E by a factor of up to almost 4.
The second rule applies to twin dosimeter situations. The NCRP [17] recommends the Rosenstein-Webster [18] algorithm:
|
|
|
|
A study by Mateya et al [20] reveals that two-dosimeter algorithms underestimate experimentally determined effective dose values from low-energy X-rays by a factor of 1.4 to 3.3. They conclude that accurate estimation of effective dose from personal dosimeters under conditions as described here remains problematic.
With a slight adaptation of the Adam phantom to smooth the transition from chest to neck, which Ulanovsky et al [21] have shown to be feasible, MC calculation can be of advantage for solving such problems, i.e. by helping to clarify the relationship between effective dose and personal dosimeter readings for cases like those described above.
|
Conclusions |
|---|
|
TopAbstractIntroductionMethodsResults and discussionConclusionsReferences |
|---|
To generate appropriate scattered radiation for exposure of the cardiologist, the complete patient phantom is required. Using a simplified PMMA phantom of equal external shape is not sufficient.
Using a broad unidirectional beam to expose the cardiologist rather than radiation actually scattering off the patient yields overestimation of occupational effective dose and underestimation of the shielding offered by protective clothing.
The effectiveness of various types and thickness of protective clothing has been evaluated for one view of one catheterization procedure. The results of the calculations are supported reasonably well by experimental studies from the literature, although the conditions are not always exactly comparable.
The derivation of occupational effective dose from single or multiple dosimeter readings after interventional radiology and cardiac catheterizations requires further optimization, e.g. through MC simulation studies, to improve the accuracy of the estimates.
|
Footnotes |
|---|
Received for publication October 10, 2002. Revision received May 6, 2003. Accepted for publication May 30, 2003.
|
References |
|---|
|
TopAbstractIntroductionMethodsResults and discussionConclusionsReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A KARAMBATSAKIDOU, B SAHLGREN, B HANSSON, M LIDEGRAN, and A FRANSSON Effective dose conversion factors in paediatric interventional cardiology Br. J. Radiol., September 1, 2009; 82(981): 748 - 755. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Tsapaki, S. Kottou, S. Korniotis, N. Nikolaki, S. Rammos, and S. C. Apostolopoulou RADIATION DOSES IN PAEDIATRIC INTERVENTIONAL CARDIOLOGY PROCEDURES Radiat Prot Dosimetry, January 2, 2009; (2009) ncn313v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C.B. Nauer, F. Kellner-Weldon, G. Von Allmen, D. Schaller, and J. Gralla Effective Doses from Scan Projection Radiographs of the Head: Impact of Different Scanning Practices and Comparison with Conventional Radiography AJNR Am. J. Neuroradiol., January 1, 2009; 30(1): 155 - 159. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. J. Irving, G. J. Maree, E. R. Hering, and T. S. Douglas Radiation dose from a linear slit scanning X-ray machine with full-body imaging capabilities Radiat Prot Dosimetry, July 1, 2008; 130(4): 482 - 489. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. W. Schultz and J. Zoetelief DOSEMETER READINGS AND EFFECTIVE DOSE TO THE CARDIOLOGIST WITH PROTECTIVE CLOTHING IN A SIMULATED INTERVENTIONAL PROCEDURE Radiat Prot Dosimetry, April 3, 2008; (2008) ncn085v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Smans, E. Vano, R. Sanchez, F. W. Schultz, J. Zoetelief, T. Kiljunen, C. Maccia, H. Jarvinen, R. Bly, A. Kosunen, et al. Results of a European survey on patient doses in paediatric radiology Radiat Prot Dosimetry, March 1, 2008; 129(1-3): 204 - 210. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. H. Park, J. K. Lee, and C. Lee Dose conversion coefficients calculated using tomographic phantom, KTMAN-2, for X-ray examination of cardiac catheterisation Radiat Prot Dosimetry, February 1, 2008; 128(3): 351 - 358. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T Siiskonen, M Tapiovaara, A Kosunen, M Lehtinen, and E Vartiainen Monte Carlo simulations of occupational radiation doses in interventional radiology Br. J. Radiol., June 1, 2007; 80(954): 460 - 468. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R R Bridcut, E Murphy, A Workman, P Flynn, and R J Winder Patient dose from 3D rotational neurovascular studies Br. J. Radiol., May 1, 2007; 80(953): 362 - 366. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D G W Onnasch, F K Schroder, G Fischer, and H-H Kramer Diagnostic reference levels and effective dose in paediatric cardiac catheterization Br. J. Radiol., March 1, 2007; 80(951): 177 - 185. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Th. M. Jansen and J. Zoetelief Dose constraints and guidance for exposure of individuals knowingly and willingly helping in the support and comfort of individuals undergoing medical exposure Radiat Prot Dosimetry, June 1, 2006; 118(3): 315 - 324. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. W. Schultz and J. Zoetelief Estimating effective dose for a cardiac catheterisation procedure with single or double personal dosemeters Radiat Prot Dosimetry, May 1, 2006; 118(2): 196 - 204. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. W. Schultz and J. Zoetelief Dose conversion coefficients for interventional procedures Radiat Prot Dosimetry, December 1, 2005; 117(1-3): 225 - 230. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. N. Kroon 3-Dimensional rotational X-ray imaging, 3D-RX: image quality and patient dose simulation for optimisation studies Radiat Prot Dosimetry, May 17, 2005; 114(1-3): 341 - 349. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| BJR | DMFR | IMAGING | ALL BIR JOURNALS |
