| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Full paper |
National Radiological Protection Board, Chilton, Didcot, Oxfordshire OX11 0RQ, UK
 |
Abstract |
|---|
 Top Abstract IntroductionMethodResultsDiscussionConclusionsReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionsReferences |
|---|
Dosimetry is an essential element of patient protection. Monte Carlo calculations in particular have already provided coefficients that facilitate the estimation of organ doses for CT scans on adult patients [7–9]. However, similar data for paediatric CT have so far been available in relation to a few voxel phantoms derived from specific patients; a baby aged 8 weeks and a child aged 7 years [10], and a 14 year-old girl [11]. Some organ dose measurements have also been reported for standard examinations and specific paediatric physical phantoms [12, 13]. More generally, Huda et al [14] have predicted significant increases in effective dose for examinations of the head and abdomen with decreasing patient size and constant exposure (mAs) setting, although these trends are derived on the broad basis of energy imparted rather than organ doses.
The existing dose coefficients for CT published by the National Radiological Protection Board (NRPB) [9] have now been supplemented by further calculations for a complete family of six geometric phantoms, representing patients of age newborn, 1 year, 5 years, 10 years, 15 years and adult, based on those of Cristy and Eckerman [15]. Calculations have been performed for three sets of CT exposure conditions modelled in the earlier work, corresponding to the Siemens DRH (Siemens Medical Systems, Forcheim, Germany), GE 9800 (General Electric Medical Systems, Milwaukee, WI) and Philips LX (Philips Medical Systems, Best, The Netherlands) scanners. This particular selection not only allows benchmarking of results with published data for the adult, but it also covers a broad range of differences in CT scanner design (beam geometry and radiation quality). Values of effective dose normalized to absorbed dose free-in-air on the axis of rotation for examinations of adults for these three scanners are widely spaced in the range of such data observed in calculations for 27 models of CT scanner [16].
Detailed results from the present calculations will be available in due course as a software report providing normalized organ doses for the irradiation of each 1 cm transverse slice of each phantom. This paper presents a broad summary of the computational methods used and an initial analysis of the trends, illustrating the likely increases in effective doses to children from CT without changes in technique to account for patient size.
|
Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionsReferences |
|---|
Here the method is used to predict energy deposition in the tissues of an anthropomorphic phantom from a rotating X-ray source. It was found that more than 100 000 particle histories are required to produce a statistically meaningful result for major organs lying within the beam of a single CT slice of 1 cm width. Advances in the power of personal computers mean that Monte Carlo simulations may now be run on the new generation of relatively inexpensive personal computers. A calculation with 250 000 histories takes approximately 5 min to execute on the Pentium III processor (Intel Corporation, Santa Clara, CA) used for these simulations.
The Monte Carlo N-Particle (MCNP) radiation transport code [17] was used in this work. Originally developed for the study of neutrons at Los Alamos for the Manhattan Project, the code has been adapted for a wider range of particles and is currently one of the most widely used general purpose Monte Carlo codes in the world. In particular, it contains a combinatorial description of geometry, which makes it suited to the geometric surfaces used in mathematical phantoms. At present, the main drawback of the code for medical applications would appear to be that it is slower to execute than other codes, such as EGS4 [18], owing to its large number of input options. Since approximately 2000 simulations were performed in this study for three CT scanners, this is obviously an important issue.
Phantom implementation
Some changes were made to the Cristy and Eckerman [15] phantoms. The thyroid gland described by Cristy and Eckerman was changed since it contained surface equations that were higher than second order, which could not be accommodated by MCNP. A thyroid composed of two cylindrical lobes with the same mass and centroid as the original was used instead. A chin was added so that the thyroid gland would be embedded at a realistic distance from the surface of the neck. The detailed heart model was simplified into muscle surrounding a volume of blood; this was because the heart does not figure explicitly in the calculation of effective dose. An oesophagus, based upon that in the adult phantom of Zankl et al [19], was added in order to calculate effective dose. It was scaled according to mass for paediatric phantoms. The trunk of the phantom was divided into six segments in order to simplify the shapes of the soft tissue cells that surround the organs. This was necessary since there is a limit to the number of surfaces that may be used to describe a cell in MCNP.
There are also differences between the adult phantom used here and that used in earlier NRPB CT dose calculations [8, 9]. This earlier work did not specifically model the oesophagus and the dose to the thymus was used in its place for the calculation of effective dose. It also assumed that the breast was composed of a 50/50 mix of fat and soft tissue. Since it is known that the fat content of breast tissue is lower in young women, it seemed more appropriate that the composition of the breast in the paediatric phantoms used here is entirely soft tissue. For consistency and convenience the same composition was used for the adult phantom.
X-ray source description for CT
CT scans were simulated by exposing a series of contiguous transverse slices of 1 cm thickness in each phantom to X-rays emitted from sources lying on a circle around the phantom in the same plane as each slice. The slices covered the entire length of the smaller phantoms and at least the regions commonly scanned from the top of the head to the knees in the larger phantoms. For each slice position, the source position is randomly sampled from a number of 1 cm long line sources, parallel to the axis of rotation of the scanner, on a circle with radius equal to the focus-to-axis distance for the scanner in question (Figure 1
). Photons are emitted normal to each line source, but unconstrained otherwise, i.e. over 360°. It follows that most source photons will not interact with the phantoms, but these photon histories add little to computational overheads since they are quickly "killed" when the photons reach a predetermined distance from the phantoms. The photons that are tracked through the phantoms essentially arise from a fan-shaped beam from each source that is perfectly collimated to the phantom and has parallel sides 1 cm apart.
|
Three scanner models were used, covering a range of design; a Siemens DRH, a GE 9800 and a Philips LX. In the Philips LX and GE 9800 scanners the X-ray beam passes through both flat and shaped filters, whereas in the Siemens DRH scanner there is only a flat filter. For the first two scanners, the shaped filters were modelled in the Monte Carlo calculations using shape and composition data from the manufacturers. The filters were situated radially inside each line source (Figure 2), and their lateral extent effectively limited to 18, the number of line sources that could be fitted around the circle describing the source geometry. For the Siemens DRH scanner, with just a simple flat filter, the number of sources that could be fitted into the circle was not limited by geometry constraints, and was chosen to be 72. A comparison was made between doses received by the adult phantom from implementations of the Siemens DRH scanner with both 18 and 72 line sources; differences between organ doses were small and within statistical uncertainties for each 1 cm slice (generally <5% for small organs and <2% for larger organs). It was thus shown that 18 sources were sufficient to approximate the continuous circular movement of the source, without significantly affecting the calculated organ doses.
|
. They were primarily chosen to cover the range of different scanner characteristics seen in an earlier study [8], which led to significant differences in the normalized organ doses previously calculated for these three scanners. The choice was particularly influenced by the need to cover the wide range of filtration that exists for CT scanners. The flat filters in these three scanners ranged from 2.7 mm aluminium (Al) equivalent for the GE 9800, to 7.4 mm for the Siemens DRH. In addition, the GE 9800 and Philips LX have shaped filters made from polytetrafluoroethylene and Al, respectively. Note that for examinations of adults it is possible to match these scanner models to more current models, based on dosimetric characteristics, by following the procedure suggested by ImPACT [21].
|
, were those specified by the manufacturers. It was noted that the resulting energy spectra were sensitive to anode angle, a parameter that is known to vary significantly between different types of X-ray tube.
It has been shown that organ doses are sensitive to the energy spectra used in the calculations [11]. Spectra produced from the IPEM catalogue [22] were therefore compared with those produced from an earlier catalogue [23] for the same tube voltages, anode angles and flat filtration, but without filtration from shaped filters. The half-value layers (HVL) and mean energies of the two sets of spectra are very similar (see Table 1).
HVLs are also available for the three scanners, which include filtration through the thin central section of the shaped filters, when present. These have been measured by the ImPACT Group at St George's Hospital, London (S Edyvean, personal communication) and are shown in the last row of Table 1. For the Siemens DRH scanner, which does not contain a shaped filter, the measured HVL is in good agreement with the calculated HVL. Measured HVLs for the other two scanners are significantly greater than calculated HVLs, indicating that the beam quality has been changed by passage through the centre of the shaped filters. Differences in the measured HVLs between scanners are narrow, with that for the GE 9800 being slightly less than for the other two scanners. However, it should be noted that, owing to the shaped filters, the HVLs of the GE and Philips scanner spectra will increase (and the photon fluence fall) off the scanner axis of rotation.
Dosimetry
For each Monte Carlo simulation, energy deposition was tallied for 37 geometric regions, which correspond to whole organs, tissue compartments or bones in the anthropomorphic phantoms. Mean organ doses for CT scans of each 1 cm slice of the phantom were normalized to the absorbed dose to air (or the air kerma) on the axis of rotation in the absence of the phantom. Results are therefore independent of X-ray tube output, which is desirable since the selection of tube current and scan time are largely at the discretion of the user and the tube output per mAs may vary markedly between scanners of the same make and model. To convert these normalized doses into absolute organ doses it is necessary to multiply them by the CTDI measured free-in-air on the axis of rotation for the particular CT scanner model and exposure conditions required.
No attempt is made to model the complex distribution of bone marrow and endosteal bone surfaces in the skeleton of the phantoms. The model assumes a uniform mix of bone and marrow in all parts of the skeleton. Dose to bone marrow is enhanced by its proximity to bone, and was calculated in a similar manner to that of Jones and Wall [24]. Bone marrow dose enhancement factors for each bone type, i, as described by Wall et al [25] as a function of photon energy, were weighted by the energy spectrum of the Siemens scanner. The resulting energy-averaged enhancement factors, hmi, are used to calculate bone marrow dose from the energy deposited in the marrow-bone mixture in the following manner. For skeletal region i the energy deposited in the red bone marrow, Emi, is related to the energy deposited in the marrow-bone mixture, Esi, by
|
|
where Wmi is the fraction of total red bone marrow mass that is in region i and (µen/
)m/(µen/)s is the ratio of mass energy absorption coefficients for red bone marrow and for the marrow-bone mixture. Mean dose to the entire bone marrow, Drbm, is then
|
|
where M is the total mass of red bone marrow in the phantom. Mass and distribution of bone marrow is taken from Cristy and Eckerman [15]. The marrow dose enhancement factor is highest for the skull, for which it is 1.34 for the newborn and 1.21 for the adult, but is more typically in the range 1.10–1.15 for the rest of the skeleton.
A similar treatment was used to calculate doses to endosteal surfaces. In this case, however, a single energy-averaged endosteal surfaces dose enhancement factor, he, was calculated for the whole skeleton from data given by Wall et al [25]. Energy deposited in endosteal surfaces in region i of the skeleton, Eei, is thus given by
|
|
where (µen/)e/(µen/)s is the ratio of mass energy absorption coefficients for endosteal surface tissues and the marrow-bone mixture. Owing to the different geometry that applies to endosteal surfaces, which are soft tissue layers (taken by the International Commission on Radiological Protection to be 10 microns thick) lying adjacent to the internal surfaces of hard bone, the enhancement factor is rather higher than for bone marrow. It ranges in value from 2.11 for the newborn to 1.85 for the adult. Mean dose to the whole endosteal surfaces, Des, is then
|
|
where mi is the mass of skeleton region i and Ms is the total mass of the skeleton.
Model validation
The family of anthropomorphic phantoms was implemented in the MCNP input file directly from the phantom descriptions furnished by Cristy and Eckerman [15], together with the modifications discussed. Rigorous quality assurance procedures were essential to ensure correct implementation of the model, since each phantom plus the CT source description contain approximately 200 surfaces plus related transformations, and up to 150 input files were created for each phantom (one for each slice). The following five steps were taken:
Variance reduction
Variance reduction is the term used to describe techniques that may be employed in Monte Carlo modelling to optimize the amount of useful information that is provided by a fixed number of particle histories. The most commonly used variance reduction methods are perhaps source biasing, where photons travelling in a certain direction from the source are given an increased probability weighting, and geometry splitting, where photons crossing specified geometrical surfaces are split into two or more photons whose probability weightings add up to that of the original photon. The ultimate purpose of these probability "games" is to ensure that the proportion of computer time that is spent following useful particle histories is maximized.
Both source biasing and geometry splitting were investigated for this problem, but it was found that neither offered any significant advantage. The scope for geometry splitting was limited owing to the fact that energy deposition is tallied over virtually the whole phantom. With respect to source biasing, it was concluded that the convenience of the adopted source description outweighed the small reduction in computing time that a biased source might produce. The adopted figure of 250 000 histories per slice gives random errors of typically less than 5% (standard error) in the doses for organs directly in the beam. Of course, random errors in the normalized organ doses for a typical scan will be significantly less than this, since the doses from more than one slice are combined.
|
Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionsReferences |
|---|
). They are taken to the nearest 1 cm slice boundary.
|
, with estimates of the statistical errors arising from the Monte Carlo method. These figures show considerable variation in normalized effective dose between scanners for the same scans, ranging up to a factor of 2 for a scan of the trunk. This is consistent with the findings of a previous study carried out for 27 scanners and an adult geometric phantom by Jones and Shrimpton [8, 9], which included the three scanners featured in this paper.
|
It should be stressed that the errors quoted in Table 3, which are well below 1% of normalized effective dose, are only the random errors generated by the Monte Carlo simulations, and they do not reflect the much larger potential uncertainties present elsewhere in the calculations, such as those in the anatomical modelling of the phantom and in the description of the photon spectrum. It is concluded from the smallness of these statistical errors that the use of 250 000 particle histories for each 1 cm slice is quite adequate for the purposes of this study.
Normalized effective doses for the four standard scans for the adult were compared under similar conditions with results from the earlier study of Jones and Shrimpton [8, 9] using the program CTDOSE [27]. As shown in Table 4, the results agree to within 10% with the most significant differences occurring for head and neck and chest scans.
|
as a function of phantom age, with age 20 years attributed to the adult phantom. It can be seen that the normalized effective dose to children relative to the adult can vary considerably between scanner models, but in all cases doses to small children are greater than doses to adults. The enhancement of the dose to the child is greatest for scans of the head and neck. Further, the enhancement is greatest for the GE 9800 scanner; therefore normalized dose to the newborn from a head and neck scan with the GE 9800 scanner is approximately 2.5 times that for the adult. Head and neck scans with the Philips and Siemens scanners produce relative normalized doses that are only marginally less; both being approximately 2.2 times the adult dose for the newborn. Apart from the case of the newborn, the ratio of paediatric to adult normalized dose appears to follow a similar trend with respect to age for all three scanners for head and neck scans, as shown in Figure 3.
|
|
|
|
. For scans of the trunk, normalized dose to the newborn ranges from approximately 2.4 times the adult normalized dose for the GE 9800 scanner to only 1.3 times for the Siemens DRH scanner. The relative normalized doses appear to vary smoothly with age, with that for the 15-year-old being virtually the same as that for the adult for the Siemens scanner, and only 15% different for the GE 9800 scanner.
The absolute value of the effective dose for a paediatric examination may be obtained by multiplying the normalized effective dose to the adult given in Table 3 by the CTDI free-in-air on the axis of rotation for the scanner, and by the ratio of the normalized paediatric dose to that of the adult (Figures 3–6). Illustrative absolute effective doses for chest scans are shown in Table 5, using typical CTDI per mAs values [28] and typical mAs values for adult patients from a national survey of CT practices in 1989 [29]. These figures indicate that the highest absolute effective dose to the adult (10.8 mSv) is delivered by the Philips LX scanner, while the highest dose to the newborn (17.1 mSv) is delivered by the GE 9800 scanner. Doses from these two scanners are higher than doses from the Siemens scanner for all ages, being approximately a factor of two higher for small children. It should be emphasized, however, that these are illustrative figures only, since mAs settings and scan limits are at the discretion of the user, and it has been assumed that the same tube voltage and mAs values as used on adults have been used on children.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionsReferences |
|---|
show a strong inverse correlation between age, and hence the size of the patient, and normalized effective dose for all the scans considered in this study. This observation is generally in agreement with results reported by Huda et al [14] that for CT exposures of a homogeneous cylindrical phantom, mean dose to the phantom normalized to air kerma also decreases as the radius of the phantom increases. The results given here, however, go further than simply establishing a relationship between mean absorbed dose and the radius of a cylindrical phantom, in that efforts have been made to reflect the differential rates of growth for the various anatomical regions of the phantoms, and indeed for the radiosensitive organs that are used to calculate effective dose. It should be appreciated that the tissue weighting factors used to calculate effective dose are taken to be independent of age [30], which may be unrealistic. Effective dose has to be combined with risk factors in order to estimate health detriment, and various authors have estimated risk factors to be significantly greater for children than adults [5, 31]. These considerations imply that the increase in radiation risk for CT scans on children compared with adults will be even greater than the increase in effective dose.
Trends between scanners
Measurements of the HVL of the X-ray beam on the scanner axis of rotation (see Table 1) suggest that there are only small differences in the "hardness" of the photon spectra from the three scanners after passage through the central region of the shaped filters. However, there is a factor of two spread in normalized effective dose (Table 3), with the Siemens DRH scanner giving the highest normalized dose and the GE 9800 scanner the lowest. The most likely cause of this difference is the presence of the shaped filter in the GE scanner, which will have the effect of reducing photon fluence in the X-ray fan beam off the scanner axis of rotation. For a head scan, for which the irradiated volume is relatively small and centred on the scanner axis of rotation, there will be a fairly uniform radiation field across the head of the child and adult phantoms for all three scanners. Hence, curves in Figure 3 for the three scanners are close to each other. However, for scans of the trunk, where the photon fluence in the X-ray fan beam will fall off significantly at greater distances from the axis of rotation for the two scanners with shaped filters, there is a greater spread between the three scanners in the normalized effective doses for child relative to adult (Figures 4–6).
Consequently, the ratio of normalized effective dose for a paediatric head and neck scan, relative to that for an adult, appears to be largely independent of the CT machine: thus the average of the three curves given in Figure 3 may, with sensible caution, be taken as a guide for most scanners. For paediatric and adult scans of the trunk, there is greater variation between scanners in the age dependence of normalized effective dose, and it may be advisable to use scanner-specific data when estimating effective doses to children.
The examples of absolute effective doses for chest scans given in Table 5 show that, when comparing scanners, the normalized effective doses given in Table 3 bear no simple relationship to the effective doses received by patients in real examinations. Indeed, the scanner that gives the highest normalized effective dose to the adult, the Siemens DRH, gives the lowest absolute effective dose for the exposure conditions assumed in Table 5. The scanner that gives the lowest normalized dose to the adult, the GE 9800, gives the highest absolute dose to the newborn for the present analysis. The values of effective dose are in the range 6–17 mSv, confirming that substantial doses are received by patients during CT examinations of the chest. Similar estimates of dose were obtained for the other three scan regions, but have not been tabulated here since the regions do not correspond so closely to those used in common clinical practice. However, it is noted that effective doses from scans of the lower trunk can be significantly higher than for the chest. For example, effective doses for scans of the entire abdomen and pelvis of the newborn could, under particular circumstances, be in excess of 35 mSv if scan protocols have not been optimized for the newborn patient.
Comparisons with previous NRPB data
Comparisons were made with the published organ dose data for CT of Jones and Shrimpton [8, 9], which were converted to effective doses for adults normalized to air kerma for the three scanners featured here. Jones and Shrimpton [8, 9] did not calculate air kerma, but rather calculated energy deposition within a 0.5 cm diameter volume of muscle, centred on the scanner axis of rotation, with the same axial length (0.5 cm) as the CT beam slice. In this study, energy deposition was tallied over the whole volume of a 10 cm long column of air with a diameter of 0.5 cm, to reflect more accurately the dimensions of the pencil ionization chambers commonly used in CT dosimetry. These two different methods of tallying energy deposition result in a difference of only approximately 2% when both are converted to air kerma.
It is thought that most of the remaining differences for head and neck scans and chest scans shown in Table 4 are accounted for by the different implementations of the oesophagus and thyroid, which affect how much of these organs are irradiated in each scan. In this study the oesophagus is modelled as a straight cylindrical tube that lies in both the head and neck region and the chest region of the phantom, while for effective dose calculations using the Jones and Shrimpton data [8, 9], the thymus was taken as a surrogate for the oesophagus and is entirely in the chest region. Hence the contribution from oesophagus to the normalized effective dose for head and neck is higher in this study in comparison with Jones and Shrimpton, while it is lower for chest. Thus for a head and neck scan with the Siemens DRH scanner, the normalized absorbed dose to the oesophagus is 12 times higher in this study than it was using data from Jones and Shrimpton, and it presently contributes approximately 10% to normalized effective dose. Correspondingly, for a chest scan, absorbed dose to the oesophagus is only 60% of that utilizing data from Jones and Shrimpton, and presently contributes approximately 10% to effective dose. The other methodological differences between the studies do not impact significantly on normalized effective dose, but can have a larger effect on individual organ doses.
To conclude, once the above effects are taken into account, the agreement with earlier NRPB data is remarkably good, which validates the descriptions of the photon fields used in the Monte Carlo simulations. Likewise, comparisons with previous NRPB data for normalized effective dose for conventional X-ray examinations show good agreement (A Khursheed, personal communication).
Comparisons with other published data
There are few published data concerning organ and effective doses for paediatric CT from measurements [12, 13] or calculation [11, 14, 32–35] and none previously for the complete family of geometric phantoms. Comparisons of dose coefficients are confounded by differences not only in the phantoms and scanner models used, but also the scan lengths assumed for standard examinations. However, the present normalized dose data are in broad agreement, with due account of likely differences in conditions of exposure, with Monte Carlo calculations for voxel paediatric phantoms by Zankl et al [32] and Caon et al [11].
Huda et al [14] have already predicted increases in normalized effective dose with decreasing patient size for typical scan protocols using a GE Advantage Hi Speed scanner. At constant tube voltage and mAs settings, the dose to a newborn patient was, according to Huda et al, larger than that to an adult by factors of 2.3 for abdomen examinations and 5.3 for head examinations. The explicit results for the newborn and adult geometric phantoms in this paper indicate corresponding factors for the three scanners considered in the range 1.4 to 2.3 for scans of the abdomen and pelvis (Figure 5), and factors of 2.2 to 2.4 for the head and neck (Figure 3). The trend in normalized dose with age for abdominal scans, shown in Figure 5, appears broadly consistent with that predicted by Huda et al [14], whereas the trend shown in Figure 3 for head and neck scans is much less pronounced than that suggested by Huda et al. It should be appreciated that there are differences in both the scan lengths and the absolute values of normalized effective dose between the study of Huda et al and the present work, which might help to explain some of this apparent discrepancy. Further analyses of these results are in progress. However, both studies clearly indicate the significant increases in patient dose, particularly to the newborn, when children are scanned using the same mAs and tube voltage settings as for adults. There is a clear need to develop size-specific protocols for paediatric CT, optimized to provide adequate image quality for the lowest dose [35–38].
|
Conclusions |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionsReferences |
|---|
Coupled with higher lifetime radiation risk factors for paediatric patients, the results of this study confirm that CT potentially poses significantly greater radiation risks to children than to adults. It is concluded that special efforts should be made in clinical practice to reduce doses to paediatric patients from CT by the use of size-specific scan protocols for optimized imaging.
Received for publication November 29, 2001. Revision received March 21, 2002. Accepted for publication April 22, 2002.
|
References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionsReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. d. Bastos, E. Y. Lee, K. J. Strauss, D. Zurakowski, D. A. Tracy, and P. M. Boiselle Motion Artifact on High-Resolution CT Images of Pediatric Patients: Comparison of Volumetric and Axial CT Methods Am. J. Roentgenol., November 1, 2009; 193(5): 1414 - 1418. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Alessio, P. E. Kinahan, V. Manchanda, V. Ghioni, L. Aldape, and M. T. Parisi Weight-Based, Low-Dose Pediatric Whole-Body PET/CT Protocols J. Nucl. Med., October 1, 2009; 50(10): 1570 - 1578. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Loeve, M. H. Lequin, M. de Bruijne, I. J. C. Hartmann, K. Gerbrands, M. van Straten, W. C. J. Hop, and H. A. W. M. Tiddens Cystic Fibrosis: Are Volumetric Ultra-Low-Dose Expiratory CT Scans Sufficient for Monitoring Related Lung Disease? Radiology, October 1, 2009; 253(1): 223 - 229. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. D. Deak, O. Langner, M. Lell, and W. A. Kalender Effects of Adaptive Section Collimation on Patient Radiation Dose in Multisection Spiral CT Radiology, July 1, 2009; 252(1): 140 - 147. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. C. Shrimpton, B. F. Wall, T. T. Yoshizumi, L. M. Hurwitz, and P. C. Goodman Effective Dose and Dose-Length Product in CT Radiology, February 1, 2009; 250(2): 604 - 605. [Full Text] [PDF]
|
|
|
|
|
|
M. J. Wan, M. Krahn, W. J. Ungar, E. Caku, L. Sung, L. S. Medina, and A. S. Doria Acute Appendicitis in Young Children: Cost-effectiveness of US versus CT in Diagnosis--A Markov Decision Analytic Model Radiology, February 1, 2009; 250(2): 378 - 386. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Gu, B. Bednarz, X. G. Xu, and S. B. Jiang Assessment of patient organ doses and effective doses using the VIP-Man adult male phantom for selected cone-beam CT imaging procedures during image guided radiation therapy Radiat Prot Dosimetry, September 1, 2008; 131(4): 431 - 443. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. H. Lee, J. M. Goo, H. J. Ye, S.-J. Ye, C. M. Park, E. J. Chun, and J.-G. Im Radiation Dose Modulation Techniques in the Multidetector CT Era: From Basics to Practice1 RadioGraphics, September 1, 2008; 28(5): 1451 - 1459. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. A. Mettler Jr, W. Huda, T. T. Yoshizumi, and M. Mahesh Effective Doses in Radiology and Diagnostic Nuclear Medicine: A Catalog Radiology, July 1, 2008; 248(1): 254 - 263. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Huda and A. Vance Patient Radiation Doses from Adult and Pediatric CT Am. J. Roentgenol., February 1, 2007; 188(2): 540 - 546. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P C Shrimpton, M C Hillier, M A Lewis, and M Dunn National survey of doses from CT in the UK: 2003 Br. J. Radiol., December 1, 2006; 79(948): 968 - 980. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. A. de Jong, J. R. Mayo, K. Golmohammadi, Y. Nakano, M. H. Lequin, H. A. W. M. Tiddens, J. Aldrich, H. O. Coxson, and D. D. Sin Estimation of Cancer Mortality Associated with Repetitive Computed Tomography Scanning Am. J. Respir. Crit. Care Med., January 15, 2006; 173(2): 199 - 203. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. S. Brody, H. A. W. M. Tiddens, R. G. Castile, H. O. Coxson, P. A. de Jong, J. Goldin, W. Huda, F. R. Long, M. McNitt-Gray, M. Rock, et al. Computed Tomography in the Evaluation of Cystic Fibrosis Lung Disease Am. J. Respir. Crit. Care Med., November 15, 2005; 172(10): 1246 - 1252. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. A. Castellano, D. R. Dance, and P. M. Evans CT dosimetry: getting the best from the adult Cristy phantom Radiat Prot Dosimetry, May 17, 2005; 114(1-3): 321 - 325. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. D. Nawfel, P. F. Judy, A. R. Schleipman, and S. G. Silverman Patient Radiation Dose at CT Urography and Conventional Urography Radiology, July 1, 2004; 232(1): 126 - 132. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| BJR | DMFR | IMAGING | ALL BIR JOURNALS |
