BJR
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
First published online August 20, 2007
British Journal of Radiology (2007) 80, 724-730
© 2007 British Institute of Radiology
doi: 10.1259/bjr/33261679
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 Moore, C S
Right arrow Articles by Saunderson, J R
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Moore, C S
Right arrow Articles by Saunderson, J R

Full paper

A method to optimize the processing algorithm of a computed radiography system for chest radiography

C S Moore, BSc, MSc 1 G P Liney, BSc, PhD 2 A W Beavis, BSc, PhD 1,3,4 and J R Saunderson, BSc, MSc 1,3

1 Radiation Physics Department, Hull & East Yorkshire Hospitals, Princess Royal Hospital, Saltshouse Road, Kingston Upon Hull HU8 9HE, 2 Centre for Magnetic Resonance Imaging, University of Hull, Hull Royal Infirmary, Kingston Upon Hull HU3 2JZ, 3 Postgraduate Medical Institute, University of Hull, Kingston Upon Hull HU6 7RX, 4 Faculty of Health and Wellbeing, Sheffield Hallam University, City Campus, Howard Street, Sheffield S1 1WB, UK

Correspondence: C S Moore Radiation Physics Department, Hull & East Yorkshire Hospitals, Princess Royal Hospital, Saltshouse Road, Kingston Upon Hull HU8 9HE, UK. E-mail: craig.moore{at}hey.nhs.uk


   Abstract
Top
 Abstract
Introduction
 Methods and materials
 Results and discussion
 Conclusions
 References
 
A test methodology using an anthropomorphic-equivalent chest phantom is described for the optimization of the Agfa computed radiography "MUSICA" processing algorithm for chest radiography. The contrast-to-noise ratio (CNR) in the lung, heart and diaphragm regions of the phantom, and the "system modulation transfer function" (sMTF) in the lung region, were measured using test tools embedded in the phantom. Using these parameters the MUSICA processing algorithm was optimized with respect to low-contrast detectability and spatial resolution. Two optimum "MUSICA parameter sets" were derived respectively for maximizing the CNR and sMTF in each region of the phantom. Further work is required to find the relative importance of low-contrast detectability and spatial resolution in chest images, from which the definitive optimum MUSICA parameter set can then be derived. Prior to this further work, a compromised optimum MUSICA parameter set was applied to a range of clinical images. A group of experienced image evaluators scored these images alongside images produced from the same radiographs using the MUSICA parameter set in clinical use at the time. The compromised optimum MUSICA parameter set was shown to produce measurably better images.


   Introduction
Top
 Abstract
 Introduction
Methods and materials
 Results and discussion
 Conclusions
 References
 
The advantages of digital over analogue imaging are leading to the gradual replacement of traditional film screen radiography with digital systems such as computed radiography (CR). The advantages of CR include digital image processing, teleradiology, electronic archiving and optimization of the image display independently of the image acquisition. Furthermore, the wider dynamic range of CR is advantageous for imaging the variety of structures with widely differing attenuating properties encountered in chest radiography [1]. In the UK, the Ionising Radiation (Medical Exposure) Regulations 2000 [2] require operators to apply optimized radiographic techniques for each examination type. In the past, optimization studies of chest radiography have been published on comparing the performance of CR with film screen [3–6], and on the optimization of X-ray beam parameters for paediatric chest imaging [7]; other published work has examined optimum tube voltage [8–13]. However, although a wealth of information exists in the literature dedicated to the optimization of the X-ray system for CR chest radiography, very little has been published on the optimization of the CR system itself, and specifically the image processing algorithms. Andriole et al [14] have reported optimization of processing parameters for the Agfa ADC 70, Fuji 9000 and AC2 systems for site-specific anatomical regions, using subjective scoring of clinical images.

Our study presents a methodology for determining the optimum post-processing algorithm settings for chest radiography using the multiscale image contrast amplification (MUSICA) image processing algorithm for Agfa CR systems (Agfa, Peissenberg, Germany) [15] and objective measurements of image quality taken from a chest phantom embedded with test tools.

Phantom used for this study
A chest phantom for the optimization of the CR system's MUSICA processing algorithm was constructed and validated. It was based upon the LucAl phantom (Standard Dosimetric/Calibration Phantom; Center for Devices and Radiological Health, Carson, CA, USA), described by Conway et al [16]. The phantom consisted of 300x300 mm plates of polymethylmethacrylate (PMMA) and 1100 alloy aluminium (Al). The overall thickness of the phantom was 267.1 mm, with a total of 4.1 mm Al, 73 mm PMMA and a 190 mm air gap. The positions and geometry of the relevant compartments were shown by Conway et al [16] to provide accurate simulation of the primary and scatter transmission through the lung field of a patient equivalent anthropomorphic chest phantom (Humanoid Systems, Carson, CA) in the diagnostic energy range.

Servomaa and Tapiovaara [17] have shown that this simple phantom provides good spectral equivalence to the Alderson-Rando male and female anthropomorphic phantoms by comparing the X-ray spectrum behind each using a germanium spectrometer, and Shrimpton et al [18] have demonstrated that the Alderson-Rando phantom can be used as a tissue equivalent phantom at diagnostic energies.

The basic LucAl phantom has previously been adapted for use in the evaluation of image quality, with the addition of test objects [19, 20]. Vassileva [21] adapted this basic LucAl design to simulate the mediastinum, heart and soft-tissue organs in the subdiaphragmal area. A 10 mm slab of PMMA was added to the central region of the phantom to simulate the broad beam attenuation properties of the upper mediastinum. Cylindrical blocks of PMMA were added to simulate the heart and the subdiaphragmal organs with thicknesses of 64 mm and 75 mm, respectively. A 5 mm thick Al strip was added to simulate the spine. Vassileva derived the appropriate thicknesses of the added anatomical regions by measuring the optical densities (ODs) of the following four clinically important regions on a set of patient radiographs: (1) the middle lung between the ribs, (2) the central mediastinum, (3) the heart and (4) the subdiaphragm. The measured OD values were converted to incident air kerma using the characteristic curve of the film screen combination used. The spectra of the X-rays transmitted through the unmodified LucAl phantom were then derived for the same X-ray beam qualities used for the patient radiographs using a computer program based on X-ray spectra data [22]. From these data, Vassileva determined the thicknesses of the added inserts required to provide the same air kerma transmission through the modified phantom as through a patient.

In our study, the design of the phantom was further modified by embedding tools to measure the contrast-to-noise ratio (CNR) in the lung, heart and diaphragm areas of the chest, and the system modulation transfer function (sMTF) in the lung region of the chest (see Figure 1Go). The process by which this modified phantom was validated is detailed in the "Methods and materials" section of this paper.


Figure 1
View larger version (145K):
[in this window]
[in a new window]

  		Figure 1. A photograph of the phantom. The PMMA anthropomorphic insert(spine, upper mediastinum, heart and subdiapragm) are visible. The MTF tool is the tungsten square positioned in the lung region, and the wax blocks are positioned in the lung, heart and diaphragm regions.

 
Chest processing algorithm used during this study
The MUSICA image processing algorithm uses a multiscale approach, whereby different frequency bandpass ranges of the same image are created. The image is decomposed and recomposed according to the Laplacian pyramid transform [23]. Selective linear or non-linear amplification of each frequency band allows manipulation in terms of contrast enhancement, dynamic range control and spatial frequency enhancement across all scales when combined to form the output image.

The image processing is influenced by the "MUSICA parameter set" comprising four user-controlled settings:

  1. Multiscale image contrast (MUSI-contrast), which determines the amount of detail contrast enhancement. Detail contrast is the magnitude of local image intensity variation.
  2. Edge enhancement, which determines the amount of edge contrast enhancement.
  3. Latitude reduction, which attenuates the larger scale intensity variations across the image in order to emphasize the medium and small-scale details.
  4. Noise reduction, which attenuates fine grain detail contrast, thus reducing noise impression in those image regions where noise is more prominent.

The work presented in this paper describes an objective methodology for the optimization of the CR processing algorithm for chest radiography and its application to an Agfa ADC Compact Plus CR system.


   Methods and materials
Top
 Abstract
 Introduction
 Methods and materials
Results and discussion
 Conclusions
 References
 
All tests were performed in a general purpose X-ray room equipped with a Philips Optimus Diagnost TH (Philips Medical Systems, Surrey, UK) ceiling-suspended X-ray system (with 3 mm Al total filtration) and an Agfa ADC Compact Plus CR reader with MD30 plates (35 x 43 cm, effective pixel pitch of 0.167 mm). In line with the current protocol used for chest imaging in our radiology department, an anti-scatter grid was not used.

Unless otherwise stated in subsequent sections, the phantom was exposed under typical conditions and with exposure factors in accordance with local practice. These were 75 kVp, 5 mAs with a focus to cassette distance of 1.80 m and with the CR plate placed 5 cm behind the phantom in the cassette holder. The X-ray field was collimated to the edges of the CR cassette. Since different CR phosphors do not have exactly matching sensitivities, a single CR cassette was used during the study. After exposure, the CR cassette was read out using the following digitizer acquisition settings as standard for chest radiography:

  1. Examination: CXR
  2. Sub exam: Chest PA
  3. Speed class: 400

These parameters define the first stage in image post processing. "Examination" is the study type, e.g. chest X-ray (CXR). This selects the histogram analysis algorithm for the relevant anatomical examination and identifies the range of useful pixel values within the defined collimated area. "Sub exam" defines the exposure view, e.g. AP, PA, LAT etc. "Speed class" sets the amount of amplification provided by the photomultiplier tube in the CR digitizer.

Software tool
An in-house analysis tool was written in Matlab (The MathWorks, Natick, MA) to read in all DICOM images directly from the Agfa CR system and measure and calculate the CNR and sMTF.

Optimization of the CR system was subsequently carried out using data derived from this analysis tool.

Validation of the phantom for the optimization technique
Although the Vassileva modified phantom has been validated previously [21], it was felt prudent to validate the phantom used in this study given its additional modifications. Five patient radiographs were acquired with exposure factors of 75 kVp and 5 mAs (i.e. the X-ray beam energy in which this experimental work was performed). These were analysed and the mean pixel values in the lung, heart and diaphragm regions were compared with the mean pixel values from the corresponding representative regions of the phantom. The mean pixel values were converted into CR cassette incident air kerma using a derived relationship between pixel value and incident air kerma, found using the following methodology.

An ion chamber was attached to the CR cassette behind the lung, heart and diaphragm region of the phantom. A number of exposures were made at a range of mAs values. These were selected to give measured air kerma values at the cassette between approximately 0.5 µGy and 30 µGy. After each exposure, the cassette was read using the standard CR acquisition parameters described above and the image was transferred to the DICOM server. The relationship between the mean pixel value and incident air kerma in each phantom region was evaluated. A relationship was subsequently derived for each area of the phantom. Finally, the mean incident air kerma values through each region of the phantom and the patients were compared.

A further validation measure was undertaken using an index calculated by the Agfa system. For each image, the histogram of pixel values is computed, and a measure of dose to the CR cassette is determined as the logarithm (base 10) of the median pixel value, and is referred to as the lgM. It is dependent on exposure (unattenuated, attenuated and scattered radiation) to the CR plate and the CR acquisition settings. The phantom was set up, exposed as described and the cassette was read using the standard CR acquisition settings. The mean lgM of 10 phantom images was compared with that of 10 clinical images.

Image quality measurements: low-contrast detectability
To assess low-contrast objects (such as subtle lung nodules, small tumours with diffuse edges and pulmonary infiltrates), paraffin wax blocks were embedded in the lung, heart and diaphragm regions of the phantom. Paraffin wax was chosen for its tissue equivalence and availability [24].

The CNR of the wax blocks in each phantom compartment was chosen as the figure of merit for the analysis of the optimum MUSICA parameter set for low-contrast detectability as it was easy to implement and we wanted to observe the effect changing the MUSICA parameter set had on the CNR of a tissue equivalent material (i.e. paraffin wax) relative to its background. The CNR was computed using the following equation:


Formula 001

where Formula is the average value of the pixels in a region of interest (ROI) within the wax block or background, and {sigma}wax/bgd their standard deviation.

The phantom was exposed using the experimental set up described and the CR cassette was read out immediately using the standard CR acquisition settings. A radiograph of the phantom with the wax blocks is shown in Figure 2Go.


Figure 2
View larger version (103K):
[in this window]
[in a new window]

  		Figure 2. A radiograph of the chest phantom complete with MTF tool in the lung region for measuring limiting spatial resolution and wax blocks in the lung, heart and diaphragm region for measuring contrast resolution.

 
Each image was sent to the clinical workstation, and multiple copies of the image, each with a different set of MUSICA parameters, were sent to the DICOM server for analysis.

Each MUSICA parameter can be altered from a value of 0 to 5, giving a potential number of combinations of 64 = 1296. Analysis of this number of images was deemed impractical, so an iterative approach was adopted, which is demonstrated in the flowchart presented in Figure 3Go. This procedure was followed until the optimum MUSICA parameter set was found. Optimization was determined after the analysis of around 600 manipulated images of the same radiograph.


Figure 3
View larger version (26K):
[in this window]
[in a new window]

  		Figure 3. Flow chart depicting the iterative methodology used to optimize CNR.

 
Image quality measurements: spatial resolution
Having optimized for contrast, the effect of adjusting the MUSICA parameter set on the spatial resolution of the system was studied. The modulation transfer function (MTF) is widely used as the metric of choice for measurement of the resolution properties of radiographic systems [25], and so the "system" Modulation Transfer Function (sMTF) in the lung area of the phantom using an edge device was chosen as the figure of merit for this part of the study. Measurement of the MTF was based on work reported by Samei et al [26], who measured the pre-sampled (detector) MTF with an edge test tool. This gives the MTF prior to any clinical digitizing or processing (hence the term pre-sampled). In this study, we are interested in the spatial resolution of post-processed images, and therefore we report changes in the system MTF (the MTF after clinical processing). The edge device consisted of a 5 x 5 cm2 tungsten plate with a thickness of 1 mm and 99.95% purity. One edge of the device was polished to ±5 µm smoothness. The edge device was placed at an angle of approximately 3° to the vertical within the lung region of the phantom, which is consistent with the recommended 1–6° [26] to minimize aliasing. The phantom was set up as previously described and exposed using 75 kVp and the required mAs to deliver approximately 60 µGy [26] to the CR receptor behind the lung region. Lower doses make it increasingly difficult to enable the line spread function (LSF) of the edge tool to be discerned over background noise, leading to a poor MTF calculation. As we are looking at relative changes only in the MTF, the use of doses greater than those used clinically was considered justified. The image plate was read using the standard CR acquisition settings and the image was sent to the clinical processing workstation. MUSICA settings were altered iteratively, starting from the optimum MUSICA parameter set determined from the CNR investigation. Following the standard methodology [26], the system MTF was calculated in both pixel directions in order to obtain the mean system MTF for both the scan and sub-scan directions of the CR plate. The sMTF at the 10% level (sMTF10%) was chosen as the optimization index, as this is a measure of the limiting spatial resolution [27]. Therefore, a MUSICA parameter set giving maximum limiting resolution was determined using this index.


   Results and discussion
Top
 Abstract
 Introduction
 Methods and materials
 Results and discussion
Conclusions
 References
 
Validation of the phantom for the optimization technique
The relationship between pixel value and CR cassette incident air kerma (kincident) for each region of the phantom was determined to be:


Formula 003



Formula 004



Formula 005

Figure 4Go shows mean CR receptor kerma calculated using Equations 2–4 through each region of the phantom and patient at 75 kVp/5 mAs.


Figure 4
View larger version (16K):
[in this window]
[in a new window]

  		Figure 4. Mean kerma from primary and scatter fluence to the CR receptor beyond the lung(dotted bars), heart (white bars) and diaphragm (striped bars) regions of the phantom and patient at an X-ray beam quality of 75 kVp 5 mAs–1. The bar on the left of each corresponding chest area represents the phantom measurement; the bar on the right represents the patient measurement.

 
The results demonstrate that, within the error limits (2SD), the air kerma behind each region of the phantom is equivalent to the mean air kerma calculated for each corresponding patient anatomical region.

Further validation is provided by analysing the lgM calculated by the Agfa system when the patient/phantom images were processed using the standard acquisition settings.

The mean lgM (±2 SD) of 10 patient images is as follows:


Formula 006

The mean lgM (± 2 SD) of 10 phantom images is as follows:


Formula 007

This suggests the total dose to the CR cassette due to scatter, unattenuated and attenuated radiation is similar for the phantom and patient.

These results, together with previously published work [17, 21], indicated that the phantom is validated as a clinical equivalent for the purposes outlined in this paper.

Image quality measurements: low-contrast detectability
Based on the provision of maximum CNR values in the lung, heart and diaphragm, the optimum MUSICA parameter set (for the beam parameters and CR acquisition setting used) was determined and is shown in Table 1Go.


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

  		Table 1. Optimum MUSICA parameter set based on maximum CNR of paraffin wax blocks alone

 
Image quality measurements: spatial resolution
Table 2Go shows a sample of the results used to find the optimum MUSICA parameter set for maximizing the sMTF10%. Each MUSICA parameter was individually adjusted from 0 to 5 while keeping all others constant at the values providing maximum CNR.


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

  		Table 2. Effect of adjusting each MUSICA setting on the sMTF10%. Each MUSICA setting is individually altered from a value of 0 to 5 whilst keeping all others constant. The affect on the sMTF10% is shown for 0 and 5 only (i.e. the extreme values)

 
The results in Table 2Go show that the sMTF10% is insensitive to latitude reduction and noise reduction. Therefore, the optimum MUSICA setting value for those parameters is identical to those quoted for the CNR measurements in Table 1Go. However, sMTF10% is sensitive to the edge enhancement and MUSI contrast parameters, and was found to be minimum at 0 and maximum at 5. Therefore, based on maximum sMTF10%, the optimum MUSICA parameter set for highest limiting spatial resolution is shown in Table 3Go.


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

  		Table 3. Optimum MUSICA parameter set for clinical chest radiography based on maximum sMTF10%

 
It is clear there is a conflict between maximum CNR and maximum sMTF10% with the MUSI contrast and edge enhancement parameters. Further work to determine the importance of low-contrast detectability relative to spatial resolution for chest radiography would need to be undertaken before definitive values of these two parameters could be recommended.

The current clinical settings for the MUSICA parameters used for chest imaging at the hospital in which this work was undertaken did not match any of the optimum settings for either low-contrast detectability or spatial resolution. Therefore, a study was undertaken to determine whether clinical image quality could be improved using a compromised optimum MUSICA parameter set based upon the work reported here. The latitude reduction and noise reduction parameters were set to the optimum values for CNR. The two parameters that manipulate both CNR and sMTF10% (the MUSI contrast and edge enhancement parameters) were set half way between their respective optimum values. Two reporting radiologists and one reporting radiographer each scored two sets of 10 clinical chest radiographs, one set processed with the compromised optimum MUSICA parameter set, the other with the current clinically used set (all parameters set to 2). Each chest radiograph was taken with typical exposure factors of the radiology department and were presented to the reviewers randomly and blindly. Image assessment criteria were based on the European Guidelines on Quality Criteria for Diagnostic Radiology Images [11], being bone detail, soft tissue, sharpness and image noise. The assessors were asked to score each criterion as follows:

  1. 0–1.9 = poor
  2. 2.0–3.9 = acceptable
  3. 4.0–5.9 = very good

Table 4Go shows the average score for the assessors, which indicates that with the compromised optimum MUSICA parameter set image quality is "very good", whereas with the original set the image quality is only "acceptable". Further work on determining the correct values for the MUSI contrast and edge enhancement parameters should lead to even better image quality than that reported here for the compromised optimum MUSICA settings.


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

  		Table 4. Average score for each of the assessment criteria. Compromised optimum and original images were scored by experienced image evaluators

 

   Conclusions
Top
 Abstract
 Introduction
 Methods and materials
 Results and discussion
 Conclusions
References
 
A methodology has been successfully developed to optimise the post-processing algorithm of an Agfa CR system for chest radiography using a validated chest phantom as a clinical equivalent. The contrast-to-noise ratio and system modulation transfer functions were used as figures of merits for low-contrast detectability and spatial resolution, respectively.

The study has indicated two optimum MUSICA parameter sets: one for maximum low-contrast detectability and one for maximum limiting spatial resolution. Further experimental work and more refined clinical trials are required in order to determine the importance of low-contrast detectability relative to spatial resolution in chest radiography, and hence the definitive optimum MUSICA parameter set. Nevertheless, a compromised optimum MUSICA parameter set was shown to produce a significant improvement in image quality when compared with the default set previously used at one general hospital.

Although the method outlined in this paper is robust, it has one limitation in that the optimization process for spatial resolution (MTF) acquired images was not produced under clinical conditions (i.e. high receptor dose). Future research should aim to overcome this limitation.

We feel that this methodology could be used by other departments using digital image acquisition to optimize their own systems using their specific clinical set-up.


   Acknowledgments
 
We are grateful to Mr Rob Walker and his staff in the Medical Physics Workshop at the Princess Royal Hospital for construction of the phantom. Many thanks go to Alistair Mackenzie of KCARE (King's Centre for the Assessment of Radiological Equipment) for independently validating the software tool. We would also like to thank Debbie Cook and her radiographer colleagues for their patience during our work at the Radiology Department at the Hull Royal Infirmary. Thanks also to Drs Paddon and Simpson and Mr Steve Balcam for scoring clinical images.

Received for publication April 11, 2006. Revision received January 12, 2007. Accepted for publication February 2, 2007.


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

  1. Goodman LR, Wilson CR, Foley WD. Digital radiography of the chest: problems and promises. Am J Roentgenol 1998;150:1241–52.
  2. Ionising Radiation (Medical Exposure) Regulations 2000. UK: The Stationery Office Limited, 2000
  3. Cook LT, Insana MF, McFadden MA, Hall TJ, Cox GG. Comparison of low contrast detectability of a screen film system and third generation computed radiography. Med Phys 1994;21:691–5.[CrossRef][Medline]
  4. Ganten M, Radeleff B, Kampschulte A, Daniels MD, Kauffmann G, et al. Comparing image quality of flat panel chest radiography with storage phosphor radiography and film screen radiography. Am J Roentgenol 2003;181:171–6.[Abstract/Free Full Text]
  5. Schaefer-Prokop CM, Prokpop M, Schmidt A, Neitzel U, Galanski M. Selenium radiography versus storage phosphor and conventional radiography in the detection of simulation of chest lesions. Radiology 1996;201:45–50.[Abstract/Free Full Text]
  6. Bacher K, Smeets P, Bonnarens K, De Hauwere A, Verstraete K, et al. Dose reduction in patients undergoing chest imaging. Am J Roentgenol 2003;181:923–9.[Abstract/Free Full Text]
  7. Hufton AP, Doyle SM, Carty HML. Digital radiography in paediatrics: radiation dose considerations and magnitude of possible dose reduction. Br J Radiol 1998;71:186–99.[Abstract]
  8. Nobuhiro O, Hajime N, Seiichi M, Kunihiro T, Katsumi N, Akira Y. Optimal beam quality for chest radiography. Invest Radiol 1996;31:126–31.[CrossRef][Medline]
  9. Dobbins JT, Rice JJ, Beam CA, Ravin CE. Threshold perception performance with computed radiography: implications for chest radiography. Radiology 1992;183:179–87.[Abstract/Free Full Text]
  10. Honey ID, MacKenzie A, Evans DS. Investigation of optimum energies for chest imaging using film-screen and computed radiography. Br J Radiol 2005;78:442–7.
  11. Council of European Communities. European guidelines on quality criteria for diagnostic radiographic images, Rep EUR 16260 EN, 1996
  12. Tingberg A, Sjostrom D. Search for optimal tube voltage for image plate radiography. Invest Radiol 1996;31:126–31.[CrossRef][Medline]
  13. Chotas HG, Floyd CE, Dobbins JT, Ravin CE. Digital chest radiography with photostimulable phosphors: signal-to-noise ratio as a function of kilovoltage with matched exposure risk. Radiology 1993;186:395–8.[Abstract/Free Full Text]
  14. Andriole KP, Gould RG, Webb WR. Finding specific display presets for computed radiography softcopy reading. J Digit Imaging 1999;12:(2 Suppl 1)3–5.[Medline]
  15. Vuylsteke P, Shoeters E. Multi Scale Image Contrast Amplification. Proc SPIE 1994;2167:550–60.
  16. Conway BJ, Butler BF, Duff JE, Fewell TR, Gross RE, Jenning RJ, et al. Beam quality independent attenuation phantom for estimating patient exposure from X ray automatic exposure controlled chest examinations. Med Phys 1984;11:827–32.[CrossRef][Medline]
  17. Servomaa A, Tapiovaara M. Patient equivalent phantoms in chest radiography. In: Moores B, Stieve F, Eriskat H, Schibilla H, editors. Technical and Physical Parameters for Quality Assurance in Medical Diagnostic Radiology: Tolerances, Limiting Values and Appropriate Measuring Methods. BIR Report 18. London: British Institute of Radiology, 1989
  18. Shrimpton PC, Wall BF, Fisher ES. The tissue-equivalence of the Alderson-Rando anthropomorphic phantom for x-rays of diagnostic qualities. Phys Med Biol 1981;26:(1)133–9.[CrossRef][Medline]
  19. Kaczmarek R, Conway B, Slayton R, Suleiman O. Results of a nationwide survey of chest radiography: comparison with results of a previous study. Radiology 2000;215:891–6.[Abstract/Free Full Text]
  20. Leitz W, Hedberg-Vikstrom B, Conway BJ, Showalter CK, Rueter FG. Assessment and comparison of chest radiography techniques in the United States and Sweden. Br J Radiol 1990;63:33–40.[Abstract/Free Full Text]
  21. Vassileva J. A phantom for dose-image quality optimization in chest radiography. Br J Radiol 2002;75:837–42.[Abstract/Free Full Text]
  22. Cranley K, Gilmore BJ, Fogarty GWA, Desponds L. Electronic version prepared by D Sutton. Catalogue of diagnostic X ray spectra and other data, The Institute of Physics and Engineering in Medicine Report No. 78. York, UK: IPEM, 1997
  23. Burt PJ, Adelson EH. The Laplacian pyramid as a compact image code. IEEE Trans on Communications 1983;34:532–40.
  24. International Commission on Radiological Units and Measurements. Tissue Substitutes in Radiation Dosimetry. ICRU Report 44. Bethesda, MD: ICRU, 1989
  25. Samei E, Ranger NT, Dobbins JT, Chen Y. Intercomparison of methods for image quality characterization. 1. Modulation transfer function. Med Phys 2006;33:(5)1454–65.[CrossRef][Medline]
  26. Samei E, Flynn MJ, Reimann DA. A method for measuring the presampled MTF of digital radiography systems using an edge test device. Med Phys 1998;25:102–13.[CrossRef][Medline]
  27. Suryanarayanan S, Karellas A, Vedantham S, Onishi SK. High resolution imager for digital mammography: physical characterization of a prototype sensor. Phys Med Biol 2005;50:3957–69.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 Br. J. Radiol.Home page
C S MOORE, J R SAUNDERSON, and A W BEAVIS
Investigating the exposure class of a computed radiography system for optimisation of physical image quality for chest radiography
Br. J. Radiol., September 1, 2009; 82(981): 705 - 710.
[Abstract] [Full Text] [PDF]

Home page
 Br. J. Radiol.Home page
C S MOORE, A W BEAVIS, and J R SAUNDERSON
Investigation of optimum X-ray beam tube voltage and filtration for chest radiography with a computed radiography system
Br. J. Radiol., October 1, 2008; 81(970): 771 - 777.
[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 Moore, C S
Right arrow Articles by Saunderson, J R
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Moore, C S
Right arrow Articles by Saunderson, J R

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