BJR
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
First published online August 17, 2006
British Journal of Radiology (2007) 80, 26-31
© 2007 British Institute of Radiology
doi: 10.1259/bjr/59538862
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 Warren-Forward, H
Right arrow Articles by Cook, A
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Warren-Forward, H
Right arrow Articles by Cook, A

Full paper

An assessment of exposure indices in computed radiography for the posterior–anterior chest and the lateral lumbar spine

H Warren-Forward, BSc, PhD L Arthur, BMRS L Hobson, BMRS R Skinner, BMRS A Watts, BMRS K Clapham, BMRS D Lou, BMRS and A Cook, BMRS

Medical Radiation Science, School of Health Sciences, Faculty of Health, University of Newcastle, University Drive, Callaghan, NSW 2308, Australia

Correspondence: Dr Helen M Warren-Forward, Medical Radiation Science, Faculty of Health, University of Newcastle, Box 16, Hunter Building, University Drive, Callaghan, Newcastle, NSW 2308, Australia. E-mail: helen.warren-forward{at}newcastle.edu.au


   Abstract
Top
 Abstract
Introduction
 Equipment and methods
 Results
 Discussion
 Conclusion
 References
 
Studies have indicated that computed radiography (CR) can increase radiation dose to the patient, leading to potential biological effects. Although manufacturers have set parameters to safeguard against overexposure, it is unclear whether these are being used by radiographers or if their recommended values are consistent with the ALARA principle. The research aims are to investigate (i) whether radiographers are producing images with exposure indices within the manufacturers recommended range (MRR); (ii) the phenomenon of exposure creep, and (iii) the relationship between exposure indices (EIs) and radiation dose. A retrospective analysis of exposure indices over an 18-month period for the posteroanterior (PA) chest and lateral (LAT) lumbar spine at two centres using Kodak 800 and 850 CR systems was conducted. A phantom study was performed to assess the relationship between EI and entrance surface dose (ESD) for fixed and varying tube potentials. Kodak recommends that images have EIs between 1700 and 1900. Thirty percent of LAT lumbar spine examinations at hospital B and 38% of PA chest examinations at hospital A were produced with EIs below 1700. In the phantom study, when using a varied tube potential (70–125 kVp) and maintaining a constant EI of 1550, ESD was reduced by 56%. All clinical and phantom images were assessed to be of a diagnostic quality. The retrospective results indicate that there is a potential to reduce the MRR and optimize patient dose. There is also evidence to suggest that EI is not a reliable indicator of patient dose. The authors recommend that staff training is essential on these newer systems.


   Introduction
Top
 Abstract
 Introduction
Equipment and methods
 Results
 Discussion
 Conclusion
 References
 
The protection of both patients and staff is essential for all radiographic examinations and it must be ensured that techniques are employed to keep radiation dose to a minimum [1]. To work within these guidelines, the radiographer needs a detailed knowledge and understanding of both the equipment they are using and how to select appropriate technical factors [2].

Since the discovery of X-rays by Roentgen in 1895, film has usually been the universal detector in diagnostic radiography. Technological advances in image acquisition techniques have been evolving over the past two decades with the introduction of the first Fuji 101 (Fujifilm; Minato-Ku, Tokyo, Japan) computed radiography (CR) system in 1983. By 1992, other manufacturers (Kodak (Australasia) Pty. Ltd, Coburg, Victoria, 3058 Australia and Agfa, Agfa Healthcare, Burwood, Victoria, 3125) entered the CR market place [3].

The main advantages of CR include: wider dynamic range (latitude), ability for post processing (manipulation of radiographic contrast and brightness), multiple viewing options, electronic transfer and storage options [4]. The wide exposure latitude of CR means that a greater range of radiation exposures (mainly current–time product (mAs)) can be used to produce a diagnostic image [4]. Underexposure and overexposure with film–screen radiography (FSR) are associated with poor image quality and the film may need to be repeated; whereas underexposure in CR results in a grainy image (high noise) that could be tolerated with some post processing, and higher than appropriate exposures in CR results in an increased image quality due to an increase in signal–noise ratio [5–8]. Gross artefacts are not evident on CR until exposure levels are 100 times lower or 500 times greater than those used in FSR [8]. Thus, while there is a potential for reduced repeat examinations and hence dose reduction in CR, it has been reported that radiographers develop a tendency to use higher than necessary exposure factors to improve image quality and avoid repeat radiographs. This tendency to increase exposure factors is referred to as exposure factor creep [6].

While FSR allows the radiographer to visually assess whether exposures are optimum, this direct visual assessment is lacking with CR systems. To allow for exposure parameter control and to safeguard from overexposure in CR, manufacturers have introduced numerical parameters that are a direct estimate of the incident exposure on the imaging plate [8]. Each manufacturer provides its own "exposure indicator", with an algorithm for its calculation; for example, this is called the exposure index (EI) by Kodak [9] and sensitivity (S) value by Fuji [10].

The EI used by Kodak represents the average raw data pixel value within a set anatomical region; a higher than average exposure is represented by an elevated EI with an increase in the EI by 300 representing a doubling of the plate exposure [8]. Kodak recommends that for all examinations, the EI should fall between 1700 and 1900 [11]. Unlike the Kodak convention, the Fuji S value decreases if the exposure increases [11], therefore the radiographer needs to have direct knowledge and training of the CR system they are using.

Even though these indices are provided by manufacturers to assist in providing a suitable image, it has been proposed that these can also be used as an indicator of patient dose levels [7, 8]. If the amount of radiation reaching the imaging plate is higher than recommended by the manufacturer, this could imply that the patient is receiving too much radiation or that the index recommended is too low (too high depending on system in use). The literature suggests that the recommended indices are not optimal and set too high, leading to increased patient dose [8].

The current research aims to investigate (i) whether radiographers are producing images with exposure indices within the manufacturers recommended range (MRR), (ii) the phenomenon of exposure creep, and (iii) the relationship between EIs and radiation dose.


   Equipment and methods
Top
 Abstract
 Introduction
 Equipment and methods
Results
 Discussion
 Conclusion
 References
 
The study comprises two parts; the first is a retrospective analysis of exposure indices for two examinations in two radiology departments. The second is an experimental design using a phantom investigating the relationship between the exposure indices and entrance–surface dose (ESD) at varying applied potentials (kVp).

Retrospective study
The Kodak Computed Radiography Systems 800 and 850 were used throughout this study. The difference between the systems is that the 800 reads fewer image plates per hour than the 850 (11). Two hospitals were selected for this study; Centre A installed CR in late 2000 and is a large teaching hospital, with a high patient throughput and a large number of radiographers. Centre B installed CR in mid 2004 and is smaller both in terms of patient throughput and radiographer numbers.

The PA chest and lateral (LAT) lumbar spine examinations were chosen as the PA chest is the most frequent examination performed (approximately 30%) in Australian radiology departments [12] and the LAT lumbar spine examination represents a high entrance surface dose examination [13]. These projections also examine different regions of the body and require different positioning and exposure factors.

Data were collected from January 2004 to June 2005 at centre A and between July 2004 and June 2005 for centre B. For the PA chest examination, 100 patients (first 10 patients (sorted alphabetically) from each day for the first 10 days) each month was sampled, while for the LAT lumbar spine (lower frequency examination) all available data was collected over the time period. Data collected included information on the examination performed, exposure index, and date and time of examination (allowing for "in or out" of hours analysis). Only data from patients over the age of 16 years were analysed.

Phantom study
The phantom study was conducted in hospital B. The phantom used was a Lung/Chest phantom (model CNR/R5330; Oxford Scientific, Silverwater, NSW 2128, Australia). The phantom was positioned as for a PA Chest examination and a series of exposures were taken at 125 kVp, 180 cm focus–film distance (FFD) under automatic exposure control, as is standard in this department. The imaging plate was read using standard chest processing algorithm and the average of the 3 EIs (1550) was used to control all other exposure factors.

The phantom was then exposed to a range of tube potentials (70–125 kVp) ensuring that the EI on the imaging plate remained fairly constant at 1550 (Table 1Go). Three high sensitivity LiF:Mg,Cu,P thermoluminescent dosemeters (TLDs) (GR-200A; Solid Dosimetric Detector and Method Laboratory, Beijing, China) were positioned on the entrance surface of the phantom for each exposure, with the mean being calculated and used in the analysis.


View this table:
[in this window]
[in a new window]

  		Table 1. Technique factors of phantom study showing relationship between entrance–surface dose (ESD) and exposure index (EI) at variable tube potential

 

   Results
Top
 Abstract
 Introduction
 Equipment and methods
 Results
Discussion
 Conclusion
 References
 
Retrospective study
Summary statistics of patient EIs for both examinations are given in Table 2Go, with the full variation demonstrated with the use of a box and whisker plot (Figure 1Go). What is immediately evident is the larger range seen in hospital A for both examinations. As the distribution of EIs is not normal, there is a need to compare the median values for each examination. Hospital B has the highest median EI for the PA chest, although it had the lowest median EI for LAT lumbar spine. The LAT lumbar spine at hospital A is the only examination that has a median EI value above the MRR of 1700 to 1900. The p-values were calculated using the Mann-Whitney statistical test for significance and indicate that the difference between the median values for each centre for each examinations was calculated to be significantly different (p<0.001).


View this table:
[in this window]
[in a new window]

  		Table 2. Summary results on exposure indices for hospital A and B for posteroanterior(PA) chest and lateral (LAT) lumbar spine examinations

 

Figure 1
View larger version (8K):
[in this window]
[in a new window]

  		Figure 1. Box and whisker plots of exposure index(EI) at hospitals A and B for (a) posteroanterior (PA) chest and (b) lateral (LAT) lumbar spine examinations.

 
More than two thirds of PA chest examinations performed at hospital B are within the MRR. However, less than one third of examinations performed at hospital A are within these limits, with the majority (38%) falling below the MRR (Figure 2aGo). For the LAT lumbar spine, approximately half of the examinations fell within the MRR for hospital B, whereas almost the same percentage is above the manufacturer's EI range for hospital A (Figure 2bGo).


Figure 2
View larger version (13K):
[in this window]
[in a new window]

  		Figure 2. Histogram showing the percentage distribution of examinations which have exposure indices(EIs) below, within and above manufacturers recommended range (MRR) for (a) posteroanterior (PA) chest examination and (b) lateral (LAT) lumbar spine examinations.

 
Figure 3Go shows the variation in EIs for all patients recorded for the LAT lumbar spine at hospital A. The line of regression indicates an incidence of exposure creep with an increase of 7.1% over the 18 month monitoring period.


Figure 3
View larger version (21K):
[in this window]
[in a new window]

  		Figure 3. Scattergram illustrating exposure creep identified for the lateral(LAT) lumbar spine in hospital A.

 
Phantom study
Figure 4Go shows the relationship between ESD and EI at a fixed 125 kVp. Unlike FSR, where there would be a linear increase in optical density and mAs, there is a log relationship in dose and EI in CR. The Kodak recommendation that for all examinations the exposure index should fall between 1700 and 1900 [11] represents a 67% variation in radiation dose. For a varied tube potential and a constant EI (of 1550), a decrease in phantom dose was noted with increasing tube potential as would be expected. All resultant images were "blindly" reviewed by four radiographers, who indicated that while the images produced at the lower tube potential had higher contrast and less noise, all images were classed as diagnostic. It is acknowledged that the assessment by radiographers is a limitation of the study. However, as they usually review the quality of images before forwarding onto the radiologist for reporting, it is assumed that the radiographers would have an understanding of what is acceptable to the radiologists.


Figure 4
View larger version (11K):
[in this window]
[in a new window]

  		Figure 4. Line plot illustrating the relationship between exposure index(EI) and entrance surface dose (ESD) for a constant tube potential.

 

   Discussion
Top
 Abstract
 Introduction
 Equipment and methods
 Results
 Discussion
Conclusion
 References
 
Overexposure or underexposure could be gauged simply in FSR. However, the wider latitude of CR allows for greater flexibility in selecting exposures for an examination, leading to possible improper radiographic exposures with underexposure and overexposure being masked [7, 8]. Additionally, with the post-manipulation ability of CR, images can be adjusted to individual viewing preferences of the radiologists who make diagnoses from the images produced without having to repeat any radiographs [14].

As mentioned previously, there is a temptation for radiographers to set higher exposures than necessary without the risk of overexposure [8]. Experience has shown that when many radiology departments changed from FSR to digital, patient doses noticeably increased. This is due to many reasons; however, most commonly it is due to radiographer knowledge that an overexposed image can be resolved with post-processing techniques [4].

Regardless of the exposure indicators set by manufacturers, the values are sensitive to segmentation algorithms, tube potential, tube filtration, delay between exposure and readout, patient positioning and FFD [7]. Thus there are expected inherent differences between each radiology department, depending on the chosen tube potential, FFD and tube filtration in use. The use of a varying tube potential affects the amount of radiation absorbed by the patient and that reaching the detector. Therefore, it cannot be assumed that there is a good correlation between exposure index and patient doses.

Hospital A has a larger range of EIs for both examinations; reasons for this trend could be due to hospital A having a larger number of patients and greater number of staff, thus allowing a greater chance for variation in radiographic technique, directly impacting on EIs. Other explanations include radiographer complacency with regard to hospital A having had CR implemented for a longer period of time, or perhaps radiographers becoming aware of exposure factors that can be used to guarantee good image quality which may be outside the MRR.

The above reasons for EI variation are based on a hospital level; however, there are a number of factors which can affect EI variation on a patient level. These include: whether or not automatic exposure control (AEC) is used and, if so, was the patient centred correctly over it; differences in patient body habitus, which would affect the exposure factors required; and for the chest exam, whether or not filters are used to reduce breast shadowing.

For the LAT lumbar spine, 30% of examinations conducted by hospital B and 38% of PA chest examinations by hospital A were produced with EIs lower than MRR. These were all judged to be clinically acceptable by the reporting radiologists. This implies that the MRR may be set too high and can be reduced to values less than 1700. In the phantom study, the EIs were assessed to be 1550; again, all images were acceptable. Since a reduction in EI by 300 represents a halving of dose, the overall reduction from the median values (1770 to 1930) to 1550 would result in significant dose savings.

The first part of the phantom study shows the relationship between EI and ESD for a fixed tube potential. For a single X-ray unit, if the same tube potential is used and no other factors are changed, EI could be used as an indicator of ESD, as suggested by Peters and Brennan [8], that "it is safe to assume that a strong correlation exists between higher exposure indices and higher patient dose", and the opposite exists for low exposure indices.

The second part of the phantom study shows the relationship between a varied tube potential, EI and dose. As tube potential increased, a decreasing mAs was used to produce the same EI (1550) for each examination. The ESD as expected also decreased for each exposure, even though the EI remained constant. Other studies [8, 15] have suggested that if EIs are above the MRR, an unnecessary level of radiation reaches the imaging plate, resulting in an increase in patient dose. The results from this current study strongly suggest that EI should not be used as an indication of ESD when comparing examinations conducted between different radiographers and institutions using different tube potentials.

In the specific case of chest radiography, there are factors which may cause an increase in patient dose with changes to CR. The use of high tube potentials has been one of the main ways that patient dose has been reduced in the past. However, the nature of the phosphors used in CR is such that they are not suited to high tube potential techniques [16, 17]. CR systems absorb low energy photons more readily than those of higher energy levels, leading to excessively noisy images [18, 19]. A change to lower tube potential technique with corresponding increase in mAs and patient absorbed dose has been reported to be necessary [20, 21] in order to maintain image quality. However, Chotas et al found that, keeping the effective dose to the patient constant, the signal to noise ratio on CR chest images was almost constant between 60 kVp and 120 kVp [22]. In this present study both departments routinely use high applied potentials (above 100 kVp), demonstrating that diagnostic images are able to be produced at higher applied potentials in CR.

As stated in a study by Willis, exposure creep is a recognized phenomenon in CR [6]. A reason suggested for this is that there is a preference by radiographers and radiologists for overexposed images, rather than the grainy, noisy appearance of underexposed images. The most pronounced change in this study was the gradual increase (7.1%) in EI (Figure 3Go) over an 18 month timeframe. As well as supporting reasons stated by Willis [6], another explanation for this exposure factor creep may be due to radiographers using higher than needed exposures so a repeat radiograph is not needed.

Comparing EIs during working hours and out of hours showed that there was little difference (for example in hospital A, 0.71% and 0.27% difference for the PA chest and LAT lumbar spine, respectively) between median exposure indices for both examinations and departments. These results do not support those reported by Peters and Brennan, who showed that there were "significantly higher exposure indices for images produced out of hours than in the normal working day" [8]. Mobile chest examinations (which were the focus of the Peters and Brennan study) have inherent exposure inconsistencies; for example, the absence of automatic exposure chambers, which would be available if performed in the radiographic department. The mobile chest examination away from the radiographic department could also be expected to place more pressure on radiographers to produce a good image, reducing the risk of repeat examinations and a return to the ward, than departmental examinations.

As radiation dose optimization is a significant concern in medical imaging, this research is aimed to increase radiographer awareness of exposure indices set by manufacturers and exposures used in the clinical setting. It is important for radiographers to understand CR concepts and not become complacent, in order to assure that dose is minimized with this technology. If radiation protection issues in digital radiography are not given careful attention, radiation exposure to patients will increase without concurrent benefits [23].

Due to the significant findings of this study and importance of reducing patient dose, it is recommended that further radiographer education and training should be implemented. Initial training should include general education about CR, explanation of EIs and how they are related to patient dose. In addition, regular quality assurance programs should be set-up to record and monitor changes in EIs over time in order to control any exposure creep. The results of this program should be fed-back to the radiographers so that they can become aware of their values and how they compare to others, allowing them to take appropriate action if needed. Manufacturers should also review the acceptable range of EIs recommended as this study has shown that phantom images produced with EIs as low as 1550 and for a range of tube potentials (70–125 kVp) (phantom study) are clinically acceptable.


   Conclusion
Top
 Abstract
 Introduction
 Equipment and methods
 Results
 Discussion
 Conclusion
References
 
Radiographers were seen to use a wide range of EIs with a total of 69% (PA chest) and 73% (LAT lumbar spine) for hospital A and 31% (PA chest) and 48% (LAT lumbar spine) for hospital B outside the MRR. A total of 25% of both examinations recorded EIs of less than 1700; all were classed as diagnostic, although these numbers were higher at individual hospitals.

The phenomenon of exposure creep was also recognized. Over an 18 month period, a 7.1% increase in EIs was noted for the LAT lumbar spine at hospital A, highlighting the importance of continual radiographer training and quality assurance programs in order to keep patient doses as low as reasonably achievable.

The phantom study demonstrated that a small increase in EI produces a large increase in entrance–surface dose. This study also demonstrated that EI cannot always be used as an indicator of changes to patient dose, but that a number of other technical factors (predominately tube potential) need to be considered. The results showed that an EI of 2000 produced at 125 kVp can deliver the same patient dose as an EI of 1700 produced at 70 kVp, where the EI difference of 300 represents a doubling of dose to the detector.

The results of the retrospective study and phantom study together recognize the possibility that the Kodak MRR of 1700–1900 can be reduced for examinations of the PA chest and LAT lumbar spine, and the authors suggest that manufactures consider reassessing their current recommended levels. A reduction of as little as 10% of the MRR (1530–1700) can result in a dose saving of nearly 50% and this study has shown that EIs in this range produce diagnostically acceptable images.

Received for publication September 21, 2005. Revision received April 17, 2006. Accepted for publication May 30, 2006.


   References
Top
 Abstract
 Introduction
 Equipment and methods
 Results
 Discussion
 Conclusion
 References
 

  1. Australian Institute of Radiography. Competency based standards for radiography and radiation therapy 1998. http://www.a-i-r.com.au/documents/AIRCompBaseSt.pdf [Retrieved March 2005]
  2. Statkiewicz-Sherer M, Visconti P, Ritenour E. Radiation protection in medical radiography. 2nd edn. St Louis, MO: Mosby, 1993
  3. Bradford C, Peppler W, Dobbins J. Performance characteristics of a Kodak computed radiography system. Med Phys 1999;26:153–75.[CrossRef][Medline]
  4. International Commission on Radiological Protection. Managing patient dose in digital radiology. Elsevier: ICRP Publication 93, 2003
  5. Lu Z, Nickoloff E, So J, Dutta A. Comparison of computed radiography and film screen combination using a contrast detail phantom. J Appl Clin Med Phys 2002;1:91–6.
  6. Willis C. Computed radiography: a higher dose? Pediatr Radiol 2002;32:745–50.[CrossRef][Medline]
  7. Seibert A. Computed radiography technology. California: Department of Radiology, Univeristy of California Davis Medical Center Sacramento, 2004
  8. Peters S, Brennan P. Digital radiography: are the manufacturers' settings too high? Optimisation of the Kodak digital radiography system with aid of the computed radiography dose index. Eur Radiol 2002;12:2381–7.[Medline]
  9. Kodak. Guidelines for acceptance testing and quality control. Kodak Directview CR 800 System and Kodak Directview CR 900 System, 2001
  10. Nakajima N, Takeo H, Ishida M, Nagata T. Fuji computed radiography: technical review No.3. In: Automatic setting functions for image density and range in the FCR system. Japan: Fuji Photo Film Co., 1995
  11. Gavin C. Kodak Health Imaging Business Relations Manager. 2005. Personal communication
  12. Department NH. Profile of the Medical Radiation Technology Workforce in NSW. Sydney: NSW Health Department, 1992
  13. Martin C, Sutton D, Sharp P. Balancing patient dose and image quality. Appl Radiat Isotopes 1998;50:3–20.
  14. Weatherburn G, Bryan S, West M. A comparison of image reject rates when using film, hard copy computed radiography and soft copy images on picture archiving and communication systems (PACS) workstations. Br J Radiol 1999;72:653–60.[Abstract]
  15. Weatherburn G, Bryan S, Davies J. Comparison of doses for bedside examinations of the chest with conventional screen film and computed radiography: results of a randomised controlled trial. Med Phys 2000;217:707–12.
  16. Huda W, Rill L, Bruner A. Relative speeds of Kodak computed radiography phosphors and screen film systems. Med Phys 1997;24:1621–8.[CrossRef][Medline]
  17. Bushberg J, Seibert J, Leidholdt E, Boon J. The essential physics of medical imaging. 2nd edn. Philadelphia, PA: Lippincott, Williams & Wilkins, 2002
  18. Tucker D, Souto M, Barnes G. Scatter in computed radiography. Radiology 1993;188:271–4.[Abstract/Free Full Text]
  19. Tucker D, Rezentes P. The relationship between pixel value and beam quality in photostimulable phosphor imaging. Med Phys 1997;24:887–93.[CrossRef][Medline]
  20. Bragg D, Murray K, Tripp D. Experiences with computed radiography: can we afford the cost? AJR Am J Roentgenol 1997;169:935–41.[Abstract/Free Full Text]
  21. Wilkinson L, Heggie J. Radiation doses from common radiographic procedures: a ten year perspective. Gold Coast: Proceedings of the 1999 Annual Conference of Engineering and the Physical Sciences in Medicine, 1999
  22. Chotas H, Floyd C, Dobbins J, Ravin C. Digital chest radiography with photostimulable storage phosphors: signal to noise ratio as a function of kilovoltage with matched exposure risk. Radiology 1993;186:395–8.[Abstract/Free Full Text]
  23. Mettler F, Ringhertz H, Vano E. Managing patient dose in digital radiography (guest editorial). Elsevier: ICRP Publication 93, 2003



This article has been cited by other articles:


Home page
 radtechHome page
N. KOWALCZYK and E. COMER
Exposure Indicator Degradation From CR Plate Processing Delays
Radiol. Technol., May 1, 2009; 80(5): 401 - 409.
[Abstract] [Full Text] [PDF]

Home page
 radtechHome page
T. L FAUBER
Exposure Variability and Image Quality in Computed Radiography
Radiol. Technol., January 1, 2009; 80(3): 209 - 215.
[Abstract] [Full Text] [PDF]

Home page
 Radiat Prot DosimetryHome page
I. A. Tsalafoutas, G. A. Blastaris, A. S. Moutsatsos, P. S. Chios, and E. P. Efstathopoulos
Correlation of image quality with exposure index and processing protocol in a computed radiography system
Radiat Prot Dosimetry, June 1, 2008; 130(2): 162 - 171.
[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 Warren-Forward, H
Right arrow Articles by Cook, A
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Warren-Forward, H
Right arrow Articles by Cook, A

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