European Journal of Radiology

Transcription

1 European Journal of Radiology 72 (2009) Contents lists available at ScienceDirect European Journal of Radiology journal homepage: Digital radiography: The balance between image quality and required radiation dose Martin Uffmann a,, Cornelia Schaefer-Prokop b a Medical University Vienna, Dept. of Radiology, Waehringer Guertel 18-20, A-1090 Vienna, Austria b Academic Medical Center (AMC), Meibergdreef 9, 1105 AZ Amsterdam, Netherlands article info abstract Article history: Received 7 May 2009 Accepted 7 May 2009 Keywords: Digital radiography Computed radiography Image quality Radiation exposure Dose indicator Although the transition from conventional screen-film imaging to digital image acquisition has been almost completed during the last couple of years, examination parameters, such as tube voltage, tube current, and filtration have been adopted from screen-film technology without further adjustments. Digital systems, however, are characterised by their flexibility: the acquisition dose can be reduced at the expense of image quality and vice versa. The imaging parameters must be optimised according to the best performance of a particular system. The traditional means of dose containment, such as positioning and collimation, are as valid for digital techniques as they were for conventional techniques. Digital techniques increasingly offer options for dose reduction. At the same time, there is a risk of substantially increasing the patient dose, possibly unawares, due to the lack of visual control. Therefore, implementation of dose indicators and dose monitoring is mandatory for digital radiography. The use of image quality classes according to the dose requirements of given clinical indications are a further step toward modern radiation protection Published by Elsevier Ireland Ltd. 1. Introduction The technological development of digital radiography (DR) and computed radiography (CR) is occurring at the same pace as that for cross-sectional imaging modalities like CT or MR imaging. The recent advent of CR detector technology, such as line scanning, dual reading, and needle-crystalline detector materials, have alleviated the differences in DR detectors with regard to dose requirements. Most modern CR and DR systems now effectively offer substantial patient dose reduction compared to screen-film radiography. Unfortunately, the reverse is also possible. There is the risk of substantially increasing the patient dose, possibly without being aware of it, or of decreasing diagnostic information because of deteriorating image quality due to inadequate image processing or suboptimal image display. In the following article, options for radiologists and physicists to assess and control dose and image quality are summarised, and strategies for optimal adjustment of digital systems are discussed. Corresponding author. Tel.: ; fax: address: Martin.uffmann@meduniwien.ac.at (M. Uffmann). 2. Digital detectors compared to conventional screen-film systems Although based on different technical principles, the digital detectors have a great deal in common with screen-film radiography: Image acquisition, image presentation, and image archiving used to be bundled on a single film sheet. With digital detectors, these main functions of radiography have been uncoupled, which is a prerequisite condition for PACS. The dynamic range of digital detectors is about 400-fold compared to film-screen (Fig. 1a, b). The inverse correlation between dose and image contrast is eliminated with digital systems. Image contrast and brightness can be optimised independently. As film blackening at higher doses does not exist with digital systems, there is the hazard of dose creep, meaning an unnoticed increase in exposure over time when using digital systems with manual tube settings. A decrease in diagnostic information through deteriorating image quality due to inadequate image processing or suboptimal image display. The latter includes the regular assessment of adequate softcopy display functions X/$ see front matter 2009 Published by Elsevier Ireland Ltd. doi: /j.ejrad

2 M. Uffmann, C. Schaefer-Prokop / European Journal of Radiology 72 (2009) Fig. 1. a, b: In a phantom series with stepwise progression of dose, the difference in dynamic range between a screen-film system (a) and a flat-panel detector (b) is illustrated (courtesy of Ulrich Neitzel, Philips Medical Systems, Hamburg). 3. Dose requirements and image quality 3.1. ALARA With increasing awareness of the need for radiation protection, a paradigm shift can be observed from the principle of image quality as good as possible to image quality as good as needed. The radiation dose to patients should be as low as reasonably achievable (ALARA) while still providing image quality adequate to enable an accurate diagnosis [1,2]. ALARA does not necessarily mean the lowest radiation dose, nor, when implemented, does it result in the least desirable radiographic image [3]. What, indeed, constitutes adequate image quality is still open for discussion for the various imaging tasks. There is a multitude of studies in the literature comparing the performance of one system with another reference system to define the amount of possible dose reduction that would still achieve an image quality equivalent to that provided by the acknowledged reference. Using this approach, it is possible to survey parameters, such as the detection of artificial lesions or the semi-quantitative assessment of subjective image impression, as a surrogate for image quality and relate these parameters to a reference of dose. To define, however, the minimum level of image quality needed to reliably make a certain type of diagnosis is much more difficult. Individually defining the minimal dose to reliably answer a specific diagnostic question in a prospective manner seems to be impossible, given the vast variety of patient-related and disease-related conditions and the workflow for radiographic examinations. Reduction of patient dose according to the ALARA principle is not only a question of selecting the right detector, but also requires the optimisation of the whole imaging chain and the selection of appropriate imaging parameters Image quality classes To implement the ALARA principle, a number of international work groups introduced the concept of image quality classes (Tables 1 and 2) [1,2,4]. Three levels of image quality (high, medium, and low) and, accordingly, three dose levels (corresponding to speed classes 400, 800, and 1600) were suggested, dependent on the demands of the diagnostic question. Such a pre-examination classification of required image quality would mean that existing

3 204 M. Uffmann, C. Schaefer-Prokop / European Journal of Radiology 72 (2009) Table 1 Image quality classes. Examples for different levels dependent on given clinical indications. Image quality High Medium Low Clinical indication Sources: [1,2,5] and - Primary bone tumour - Non-displaced fracture - Lumbar spine in two projections in patients with chronic back pain with no pointers to infection or neoplasm - Control of a known displaced fracture - Control after metal implantation for osteosynthesis - Follow-up of pneumonia in adults - Follow-up in longitudinal studies, e.g., in patients with scoliosis Table 2 Relationship between image quality classes and dose requirements of different radiography systems [4]. Image quality class High Medium Low DR (400) DR (800) DR (1600) CR (200/400) CR (400) CR (800) Film-screen system (200) Film-screen system (400) Film-screen system (800) referral guidelines would require an additional parameter, such as image quality class, although it has to be clearly stated that the responsibility for such a classification should be in the hands of the radiologists. Based on the different dose efficiency of existing digital radiographic equipment, ALARA would also mean that, for the same clinical question, different exposure parameters must be applied, dependent on the radiographic equipment used. Evaluations of phantom images showed that a medium class image quality provided by a film/screen image with a speed class of 400 was achievable by a storage phosphor system (CR systems using a powder storage phosphor plate with a single-sided read-out) with doses equivalent of speed 200 and 400 and a flat panel system (DR system using a CsI/TFT detector) at a dose equivalent of speed 1600 or 25% of the dose [5]. A desirable option to optimally manage this process in a clinical routine in the future may be a kind of examination parameter database that is integrated in the system. For various examination types and image quality classes, the database offers appropriate pre-settings of manual parameters or AEC exposures [6]. Since the responsibility for such a classification should be in the hands of the radiologists, it would require reorganisation of the workflow because, unlike CT and fluoroscopic examinations, ordinary radiographic examinations are usually performed without the radiologist s knowledge or input. The image quality class must be selected beforehand, which requires a close communication between the radiologist and his clinical counterpart Strategies for dose containment In addition to image quality classes, there are a variety of measures before and during the examination that allow a net reduction in patient dose Adaptation of technical acquisition parameters to digital equipment In general, it is important to consider that each change in one of the segments of the imaging chain (generator, tube, filtration, grid, detector, workstation, PACS, monitor, or print-out) will require adaptation of the exposure parameters, and thus, of the threshold values of the automatic exposure control system (AEC). Similarly, as radiologists change the CT acquisition protocols when changing from a 16-slice-scanner to a 64-slice-scanner, they have to adapt the exposure parameters when purchasing new CR or DR equipment. In practice, this process should be integrated in the certification process before starting to work with the new equipment Allocation of examinations to the most suitable radiographic system When radiologists have multiple systems available, knowledge of the different dose requirements of the systems should be reflected in the choice of system according to optimal suitability. Examples include the following: the CsI/TFT flat panel systems produce chest radiographs of higher quality at lower dose levels compared to an a-se/tft-based DR system; charged-coupleddevice (CCD) equipment is better suited for small field applications than chest radiography; and a dual-read-out CR system works more dose efficiently that a single-read-out CR system. The aim of this allocation should be to reduce the net dose for the entire patient group Preservation of dose-saving principles validated for conventional techniques Guidelines implemented for conventional radiography, including appropriate collimation, appropriate source-to-image distance (SID), focal spot size, and patient positioning, are as valid for digital techniques as they were for conventional techniques. This is especially important in an intensive care unit or emergency department setting where many of the above-mentioned parameters are set manually. Unfortunately, there is a tendency to handle these principles less precisely, based on the fact that digital technique is more tolerant to dose variations and offers more options to retrospectively rescue image quality by processing. In conventional screen-film radiography, inadequate images were easily identified. The majority of these images were subsequently re-taken. With digital techniques, however, image processing can compensate for acquisition errors. Under certain circumstances, the source of error is invisible on the processed radiograph. In particular, an unnecessarily large collimation can be masked by electronic shutters. Appropriate collimation of the X-ray beam is important for both radiation protection and image quality in a DR/CR setting [7]. The optimal collimation area depends on the individual patient and is the responsibility of the radiographer, taking into account the patient body size, the diagnostic question, and the individual requirements of the examination type. This becomes especially important when using systems that are not cassette-based (some CR systems, all DR systems) and that use large-area detectors. Continuous training, in-services, oversight, and feedback represent valuable tools to ensure the proper use and radiographic technique in accordance with the ALARA principle Optimisation of tube voltage and beam filtration Traditionally, the limited dynamic range of conventional film/screen systems required the use of high kilovoltage settings to penetrate high attenuating areas, such as the mediastinum, while still maintaining an acceptable level of image contrast in the lungs. In general, digital images are acquired with the same kvp settings as those used for conventional film-screen. This approach, however, was recently questioned. Digital systems have redefined the traditional relationship between tube potential and image contrast because image processing allows the contrast and density of an image to be optimised independently. A second point to challenge the traditional high kvp technique for digital systems is related to the fact that all digital detector mediums have to a different extent, depending on their absorption characteristics higher dose efficiency (DQE) at lower kvp ranges. This,

4 M. Uffmann, C. Schaefer-Prokop / European Journal of Radiology 72 (2009) theoretically, can be translated to an improved signal-to-noise. [8,9]. Most studies were carried out by physicists and were based on SNR measurements or contrast detail experiments. There are few studies that evaluate the effects of low kvp settings in a clinical environment [10,11]. They uniformly describe significantly improved image quality for posterior anterior chest radiographs acquired with 90 kvp (without an increase in effective patient dose). In a recently published study, a reduction of the tube potential proved to be beneficial for both dose reduction and image quality in lumbar spine radiography [12]. Some authors recommend additional beam filtering rather than lowering the tube voltage [13,14]. By using additional filtration to harden the beam, e.g., thin layers of copper, the entrance dose can be lowered by up to 40% for certain body parts [15]. Standard tube filtration in diagnostic radiology, as required by regulations, is 2.5 mm of the aluminium (AL) equivalent. Additional filtration in the X-ray beam can be used to remove the low energy part of the tube spectrum, which is completely absorbed in the patient without contributing to image quality. While additional filters of 1 mm AL plus 0.1 mm or 0.2 mm copper are suggested, and are common practice for paediatric radiography in many countries (parts of the European guidelines for paediatric imaging), this technique is not applicable to imaging of adults due to the considerable prolongation of exposure time (>50%), most likely in many patients beyond the recommended threshold of 20 ms [13,16] Anti-scatter grids Anti-scatter grids are generally used for areas with high absorption and a high level of scattered radiation that leads to deterioration of image quality with respect to signal-to-noise ratio and contrast. Use of an anti-scatter grid is inevitably associated with a higher acquisition dose (factor 2 or 3) compared to the same image obtained without an anti-scatter grid. Based on the fact that CR and DR systems (with the exception of slit-beam technology) represent area detectors that are vulnerable to scatter effects, general practice is that anti-scatter grids are used in applications similar to those in conventional radiography (e.g., upright chest radiography, radiographs of the spine or pelvis, etc.). The use of anti-scatter grids for bedside chest radiography varies among institutions. In favour of anti-scatter grids is the fact that the lower K-absorption edge of the detector material of both CR and DR makes it more sensitive for the absorption of low-energy scattered radiation compared to film-screen systems. On the other hand, the grid can be omitted if the tube voltage is lowered, and therefore, scatter radiation is reduced. The latter is especially applicable for small volumes (e.g., paediatric applications) [3]. With adequate image processing, a sufficient image quality is achievable for most patients, considering the diagnostic requirements in supine chest radiography (Fig. 2 a c). Nevertheless, for patients with a body mass index clearly above average, image acquisition with high voltage and use of a grid is still required Optimising image processing technology There is no doubt that the processing of digital images plays a major role in successfully performing the medical imaging task. Three aspects must be taken into account: (1) The overall impression of a typical radiograph (e.g., chest radiograph) should be similar to conventional radiography performed with screen-film systems in order to recognise the typical features acquired during radiology training and to maintain the option for universal interpretation. (2) There should be a certain consistency in image processing to determine whether something is normal, and for interpretation of studies longitudinally. (3) The modern multi-frequency processing capabilities enhance image contrast differently in different anatomic regions and differently for different image structures. In reader studies, images enhanced with elaborate processing tools are greatly preferred [17]. The question of whether a gain in image quality by processing can be utilised for dose reduction remains to be answered. As image noise can be both amplified and suppressed by processing, the link between image processing and dose requirement is evident Controlling patient dose The inverse correlation between dose and image contrast is eliminated with digital systems. Therefore, film blackening as an indicator of overexposure, no longer exists. Even a 10-fold overexposure cannot be recognised as a too black digital radiograph. In digital radiography, there is a reciprocal relationship between dose and signal-to-noise ratio: a lower acquisition dose is associated with increased image noise and vice versa. The problem is that the subjective visual assessment of image noise is quite insensitive. In addition, increased acquisition dose will be rewarded by increased signal-to-noise. Thus, images will be hardly ever rejected by the radiologist when overexposed. This may consequently lead to a small but continuous increase in the amount of acquisition dose when exposure parameters are set manually. Weatherburn et al. compared acquisition doses in 269 patients admitted to the intensive care unit [18]. Patients were randomly assigned either to CR or conventional radiographic imaging of the chest. The authors found significantly higher surface entry doses, entrance surface air kerma (median 0.21 mgy vs mgy), and higher effective doses (median msv vs msv) for CR compared to conventional radiography. This underscores the supposition that when in doubt the technologists tend to use higher acquisition levels in CR than in conventional radiography Image noise In digital radiography, image noise is inversely related to the amount of detector radiation dose. Thus, visual assessment of image noise is potentially a control mechanism. However, the human visual control mechanism is not only very subjective, but also quite insensitive. Observers complained about the noise in CR images only when they were exposed to considerably less than 50% of the appropriate level [19]. By the time a too high noise level is visualised by the radiologist, there is already an increased risk for loss of diagnostic information Speed-class system Given the very limited visual control and the arbitrarily allocable amount of dose, the question arises of how dose can be controlled in digital systems. Film/screen combinations were characterised by a limited dynamic range, and thus, required a certain (rather limited) range of radiation exposure to obtain an image of optimal diagnostic quality. This dose (also considered detector exposure) was traditionally described in speed or speed class. Multiple studies comparing the performance of film/screen and digital systems described the dose used to acquire film or digital images with speed values (Table 2). This might be understandable in the context of studies that compared the performance of the two

5 206 M. Uffmann, C. Schaefer-Prokop / European Journal of Radiology 72 (2009) Fig. 2. a c: Dose reduction by optimising radiographic technique in supine chest radiography. The old standard included a grid. Parameters were set at 125 kvp and 1.25 mas (a). Substantial dose reduction was achieved by removing the grid and lowering the tube voltage to 90 kvp at 1.0 mas (b). Further dose reduction is feasible by using a needle structured CR system (90 kvp, 0.5 mas, no grid was used) (c). radiographic techniques. However, it is important to understand that the measure of speed is, in fact, not applicable to the detector dose of digital radiographs. This is illustrated by the following. Radiographic speed is formally defined as the inverse of the exposure necessary to produce a net film density (above base and fog) of an optical density (OD) equal to 1. The density of a CR or DR image, however, is entirely arbitrary and dependent on the processing and not completely related to the dose. The film speed is primarily determined by the composition and thickness of the screen and also describes the spatial resolution performance of the film/screen system. Yet, in CR/DR units, the system resolution is determined by other factors. For example, in CR, detector (phosphor) thickness and the laser spot size are determinants of resolution, and, for both CR and DR, pixel size is an important factor. While specification of a speed (e.g., through an arbitrary index term) for a given type of screen-film combination is not directly coupled to the noise level, changing the dose in a digital system will affect the signal-to-noise ratio. Even for screen-film systems, the relation between speed number and image receptor dose is defined only for homogeneous exposures under laboratory conditions, involving specific phantoms and beam qualities. For patient images under clinical conditions, the actual air kerma at the detector may deviate from the value expected from the speed number (e.g., 2.5 Gy at speed 400). It was, therefore, recently recommended, in an editorial by Huda, that authors specify the amount of radiation incident on any digital receptor when characterising the system performance, such as the detector air kerma in Gy [20].

6 M. Uffmann, C. Schaefer-Prokop / European Journal of Radiology 72 (2009) Fig. 3. List of terms for exposure indices for various digital systems and their relationship to traditional dose measure (in Gy). In the second column, the proposal for an international standardisation is detailed (courtesy of Ulrich Neitzel, Philips Medical Systems, Hamburg) Exposure index In an attempt to give the user feedback about the actual detector dose level of a clinical image, most digital systems provide what is called an exposure index. It is important to note that the definition of this exposure index was only recently standardised [21] and for systems now in use the numbers given for the exposure index still refer to different dose quantities for different systems. In all digital systems, the exposure index is derived from signals in the acquired image itself, and thus, describes the detector dose. The mathematical definition and calibration of the exposure index is different for the different manufacturers of digital systems (Fig. 3). The exposure index (EI) value provides a composite measure of patient-related features and exposure. Different combinations of patient body constitutions and exposure can result in the same detected signal. Variation of the EI may occur due to varying imaging content (e.g., with or without pneumonia in a chest radiograph), even if the same exposure setting was used and patient entrance exposure was identical. This means that the exposure index in digital systems can only be used as a surrogate for dose management; EI does not represent an equivalent for patient entrance exposure. Interpretation of these EI values is further complicated by factors such as the collimation detection, the manufacturer-specific calibration, and the examination-specific image processing. The main purpose of the exposure index is to allow for a longitudinal (time) comparison of system operation during instillation and practice, which depends on a properly calibrated system. Furthermore, the exposure index is influenced by the accuracy of the collimation detection algorithm, the collimation, the time delay between acquisition and read-out, and the reproducibility of the exposure system. Thus, for proper interpretation, it is necessary to have an idea about the degree of variation of EI attributable to varying clinical and acquisition conditions. This range then may have a variable effect on dose. At present, a broad acceptance of the exposure index is hindered by the multitude of different vendor-specific variants (Fig. 2) Automated dose control Modern digital equipment usually provides the necessary data interface for automatic data collection and evaluation [22,23]. Integrated in a suitable RIS/PACS environment, dose control can be established in clinical routine as part of an overall quality control Fig. 4. a, b: Automated data collection and evaluation for longitudinal assessment of dose, illustrated by anterior-posterior lumbar spine radiographs (a). Even small alterations of the automated exposure control are indicated (arrow). Automated assessment of the kerma-area-product in posterior anterior chest radiographs (b). The majority of exposures are below the diagnostic reference level (red line) (Courtesy of Ulrich Neitzel, Philips Medical Systems, Hamburg). program. With longitudinal assessment of dose-relevant parameters, even small deviations can be identified (Fig. 4 a, b) Diagnostic reference dose levels Diagnostic reference levels (DRL) are defined as dose levels for typical examinations for groups of standard-sized patients or standard phantoms for broadly defined types of equipment. They are specified as entrance skin air kerma (ESAK, measured in air without backscatter) or as entrance skin dose (ESD, measured in specified material with backscatter, more commonly used in the United States). The concept of DRL was introduced by the International Commission of Radiological Protection (ICRP) in the 1990s [2]. DRL are typically set at the third quartile (75% value) of the dose distribution, derived from a suitable patient dose survey. The DRL specified are not to be exceeded with routine practice (Fig. 4b). The reference levels are periodically reviewed and, if necessary and possible, modified on the basis of knowledge of current practice. In Table 3, some examples of DRL for radiographic examinations in adults for different countries are given. Specific values have also been set for paediatric examinations in different age groups. The DRL are advisory. That is, they do not distinguish between acceptable and unacceptable practice. It should be noted that the reference levels derived from these surveys represent the state of practice and not the state of the art, and should be considered as such. They do not take into account the options provided by most

7 208 M. Uffmann, C. Schaefer-Prokop / European Journal of Radiology 72 (2009) Table 3 Examples of diagnostic reference levels (ESD = entrance surface dose, ESAK = entrance-surface air kerma). US 1999 ESAK [mgy] UK 2000 ESD [mgy] AP/PA skull 3 5 Lat skull PA chest Lat chest AP abdomen AP pelvis 4 10 AP C-spine 1.25 AP L-spine Lat L-spine Sources: [22,24,25]. Germany 2003 ESD [mgy] modern CR and DR detector technologies, and allow for flexibility to accommodate individual variation in dose reduction and resultant image quality. However, because digital systems have this greater freedom in setting the dose level without overexposing, adherence to reference levels is even more important to avoid dose levels to the patient that do not contribute to the clinical purpose of a medical imaging task Quality assurance National regulations exist for the approval and operation of radiography equipment to assure that sufficient image quality is achieved with a reasonable amount of radiation dose. Technical details are specified in national or international standards. The image quality of the imaging systems is verified with technical phantoms at regular intervals. A typical phantom for this purpose comes with a copper step wedge to control the dynamic range, different low-contrast objects for assessment of the low contrast resolution, and a lead bar pattern for determination of the spatial resolution Perspective The advances in digital radiography have revolutionised the way radiologists interpret medical images and communicate the findings. Modern image processing algorithms provide optimised image quality at a reasonable dose level. Beyond this, a variety of elaborate post-processing techniques have evolved, which increasingly assist the radiologist in the differentiation between anatomic noise and abnormal findings. These diagnostic enhancements, such as dual energy subtraction, temporary subtraction, digital tomosynthesis, and computer-aided diagnosis, are described in detail in other articles of this issue. 4. Summary The balance between the necessary image quality and the amount of radiation dose invested is much more flexible for digital detectors compared to screen-film systems. Digital detectors are based on different technical principles, and therefore, require different levels of dose. However, the dose requirement differences between DR and CR are shrinking. Optimising image quality and lowering the dose requires the optimisation of the whole imaging chain (detector, acquisition, processing, and display). Guidelines implemented for conventional radiography, including appropriate collimation, appropriate source-to-image distance (SID), focal spot size, and patient positioning, are as valid for digital techniques as they were for conventional techniques. In contrast to options for dose reduction with digital systems, there is the hazard of dose creep, meaning an unnoticed increase in exposure with time. This is most problematic in portable examinations using manual tube settings, paediatric applications, and fluoroscopic studies. With longitudinal assessment of dose-relevant parameters, dose control can be established in the clinical routine as part of an overall quality control program. As part of such quality control, diagnostic reference levels should be maintained. In addition, implementing image quality classes represents active dose containment. However, this requires reorganisation of the routine workflow. References [1] Managing patient dose in digital radiology. ICRP Publication 93. Elsevier: ICRP; [2] 1990 Recommendation of the international commission on radiological protection users edition. ICRP Publication 60. Oxford: ICRP: Pergamon Press; [3] Seibert JA. Tradeoffs between image quality and dose. Pediatr Radiol 2004;34(suppl. 3):S [discussion S ]. [4] Busch HP, Busch S, Decker C, Schilz C. Image quality and exposure dose in digital projection radiography. Rofo 2003;175:32 7. [5] Busch HP, Faulkner K. Image quality and dose management in digital radiography: a new paradigm for optimisation. Radiat Prot Dosimetry 2005;117: [6] Neitzel U. Management of pediatric radiation dose using Philips digital radiography. Pediatr Radiol 2004;34(suppl. 3):S [discussion S ]. [7] MacMahon H. Digital chest radiography: practical issues. J Thorac Imaging 2003;18: [8] Launders J, Cowen A, Bury R, Hawkridge P. Towards image quality, beam energy and effective dose optimisation in digital thoracic radiography. Eur Radiol 2001;11: [9] Honey ID, Mackenzie A, Evans DS. Investigation of optimum energies for chest imaging using film-screen and computed radiography. Br J Radiol 2005;78: [10] Metz S, Damoser P, Hollweck R, et al. Chest radiography with a digital flat-panel detector: experimental receiver operating characteristic analysis. Radiology 2005;234: [11] Uffmann M, Neitzel U, Prokop M, et al. Flat-panel-detector chest radiography: effect of tube voltage on image quality. Radiology 2005;235: [12] Geijer H, Norrman E, Persliden J. Optimizing the tube potential for lumbar spine radiography with a flat-panel digital detector. Br J Radiol 2009;82:62 8. [13] Samei E, Dobbins III JT, Lo JY, Tornai MP. A framework for optimising the radiographic technique in digital X-ray imaging. Radiat Prot Dosimetry 2005;114: [14] Hamer OW, Volk M, Zorger N, et al. Contrast-detail phantom study for X-ray spectrum optimization regarding chest radiography using a cesium iodideamorphous silicon flat-panel detector. Invest Radiol 2004;39: [15] Schaetzing R. Management of pediatric radiation dose using Agfa computed radiography. Pediatr Radiol 2004;34(suppl. 3):S [discussion S ]. [16] Dobbins III JT, Samei E, Chotas HG, et al. Chest radiography: optimization of x-ray spectrum for cesium iodide-amorphous silicon flat-panel detector. Radiology 2003;226: [17] Niemann T, Reisinger C, Rau P, Schwarz J, Ruis-Lopez L, Bongartz G. Image quality in conventional chest radiography. Evaluation using the postprocessing tool Diamond View ((R)). Eur J Radiol [18] Weatherburn GC, Bryan S, Davies JG. Comparison of doses for bedside examinations of the chest with conventional screen-film and computed radiography: results of a randomized controlled trial. Radiology 2000;217: [19] Willis CE. Computed radiography: a higher dose? Pediatr Radiol 2002;32: [discussion ]. [20] Huda W. The current concept of speed should not be used to describe digital imaging systems. Radiology 2005;234: [21] Medical electrical equipment exposure index of digital X-ray imaging systems Part 1: definitions and requirements for general radiography. Geneva, Switzerland: International Electrotechnical Commission (IEC), International Standard IEC : ; [22] Schuncke A, Neitzel U. Retrospective patient dose analysis of a digital radiography system in routine clinical use. Radiat Prot Dosimetry 2005;114: [23] Ward M, Hughes D, Connolly P, Moores BM. Central dose data management and analysis in IT-driven radiation protection strategies. Radiat Prot Dosimetry 2005;114: [24] Gray JE, Archer BR, Butler PF, et al. Reference values for diagnostic radiology: application and impact. Radiology 2005;235: [25] Wall BF. Diagnostic reference levels in the X-ray department. Eur Radiol Syllabus 2004;14:66 73.