Guest Editorial, Preface, Main Points, Glossary and Chapter 1
Article Outline
- 1. Introduction
- 1.1. Purpose of this document
- 1.2. Introduction
- 1.3. History
- 1.4. What is digital radiology?
- 1.5. Differences between intrinsic digital modalities and digital radiology
- 1.6. Digitised conventional films
- 1.7. Plain radiography
- 1.8. Digital fluoroscopy
- 1.9. Comparison of digital and non-digital techniques
- 1.10. Image quality and exposure levels in digital radiology
- 1.11. Further aspects of digital radiology
- Copyright
Digital radiology may represent the greatest technological advance in medical imaging over the last decade, yet many people remain unaware of its existence. With the advent of faster computers, larger storage capabilities, and new x-ray detector systems, film for x-ray imaging is becoming obsolete. An appropriate analogy that is easy for most people to understand is the replacement of typical film cameras with digital cameras. Images can be taken, immediately examined, deleted, corrected, and cropped, and subsequently sent to a network of computers.
In many radiology departments in developed countries, film is simply no longer used in production of x-ray images. The benefits are enormous. The referring physician can often view the requested image on a desktop personal computer, usually accompanied by the interpretation, just minutes after the examination was performed. The images are no longer held in a single location; they can be seen simultaneously by physicians who are kilometres apart. In addition, the patient can have all his or her x rays on a compact disk to take to another physician or hospital.
Digital technology has the potential to reduce patient doses. This is important as medical uses of ionising radiation now represent over 95% of all man-made radiation exposures and is the largest single radiation source after natural background radiation. What then is the problem and why did ICRP Committee 3 request a Task Group to write this document?
While digital techniques have the potential to reduce patient doses, they also have the potential to significantly increase them. Experience has shown that as many radiology departments have transitioned to digital equipment, patient doses have not reduced but have increased measurably. The reasons for this are multiple. Technologists know that an overexposed image can be resolved with the computer, but an underexposed image will need to be repeated. As a result, there is a tendency to give more dose than is necessary. Most systems do not easily track unsatisfactory images that have been deleted from the system, and although the data is present, few systems display meaningful dose or exposure factors for the patient record. Recent studies also show that even if exposures are performed correctly, there is an increase in the number of examinations requested by the referring physicians, probably because of the ease and convenience of getting the images and results.
These are but a few of the issues addressed in this report. It is clear that attention needs to be given to many factors when transitioning to digital systems. A few examples include standardisation of equipment, training, privacy and security concerns, quality control, diagnostic reference levels, archiving, compression algorithms, and referral criteria.
Overall, this is a technology that is advancing rapidly and which will soon affect hundreds of millions of patients. If careful attention is not paid to the radiation protection issues of digital radiology, medical exposure of patients will increase significantly and without concurrent benefit. Conversely, if the radiation protection issues are addressed adequately, medical exposures may decrease without decreasing the diagnostic benefit to the patient.
Preface
Over the years, the International Commission on Radiological Protection (ICRP), referred to below as `the Commission', has issued many reports providing advice on radiological protection and safety in medicine. ICRP Publication 73 is a general overview of this area. These reports summarise the general principles of radiation protection, and provide advice on the application of these principles to the various uses of ionising radiation in medicine and biomedical research.
Most of these reports are of a general nature, and the Commission wishes to address some specific situations where difficulties have been observed. Reports on such problem areas should be written in a style that is accessible to those who may be directly concerned in their daily work, and every effort should be made to ensure wide circulation of such reports.
A series of such reports was initiated at the Commission's meeting in Oxford, UK in September 1997. On the recommendation of ICRP Committee 3, the Commission has established a number of Task Groups to produce reports on topical issues in medical radiation protection.
Some such reports have already appeared in print as ICRP Publications 84, 85, 86, and 87. The present report continues this series of concise and focused documents, and several more advisory reports are being prepared.
The Task Group on Dose Management in Digital Radiology was launched at the Commission's meeting in The Hague, the Netherlands, in September 2001. Its terms of reference were to review methods that are specific to digital radiography that could lead to increased radiation doses. Furthermore, the Task Group was requested to provide recommendations to manufacturers and users that can reduce potential unnecessary doses.
The membership of the Task Group was as follows:
| E. Vañó (Chairman) | B. Archer | K. Faulkner |
| B. Geiger | R. Loose |
The corresponding members were:
| B. Bergh | H.P. Busch | J.M. Fernández |
| R. Gagne | M. Rosenstein | C. Sharp |
| M. Wucherer |
The membership of Committee 3 during the period of preparation of this report was:
| F.A. Mettler, Jr. (Chairman) | J.-M. Cosset | C. Cousins |
| M.J. Guiberteau | I. Gusev | L.K. Harding (Secretary) |
| M. Hiraoka | J. Liniecki (Vice-Chairman) | S. Mattsson |
| P. Ortiz-Lopez | L.V. Pinillos-Ashton | M.M. Rehani |
| H. Ringertz | M. Rosenstein | C. Sharp |
| E. Vañó | W. Yin |
This report aims to serve the purposes described above. In order to be as useful as possible for those purposes, its style differs in a few respects from the usual style of the Annals of the ICRP.
The report was approved for publication by the Commission in November 2003.
ICRP Publication 93
Main points
Glossary and acronyms
Dictionaries of digital terms
http://www4.gartner.com/6_help/glossary/GlossaryR.jsphttp://www.highend3d.com/rosetta/http://www.learndigital.net/glossary.htm
The conversion of analogue signals into digital data – normally for subsequent use in a digital machine.
Aliasing
The Nyquist condition (see Spatial resolution limit) is not fulfilled if the pixel spacing is too large with respect to the spatial frequency content of the image. The high-frequency components beyond the spatial resolution limit will appear as artificial low-frequency components (`aliasing artefacts') in the image. The practical meaning is that smaller objects will appear artificially enlarged. There is no way to distinguish between a true large detail and an aliased small detail in the final image.
The maximum number of different pixel values is often expressed in powers of 2, which is identical to the number of bits necessary to store the pixel values (binary coding system), e.g. 4096 different pixel values require a bit depth of 12 because 212=4096. Note: The bit depth determines the contrast resolution in connection with the dynamic range.
A photo-electric device that converts light information into electronic information. CCDs are commonly used in television cameras and image scanners, and consist of an array of sensors that collect and store light as a build-up of electrical charge. The resulting electrical signal can be converted into computer code and reconstructed to form an image (see the ACR standards, available at http://www.acr.org/dyna/?doc=departments/stand_accred/standards/standards.html).
Used for evaluation of digital systems. It is possible to test the image quality and the observer's perception. The phantom consists of a Plexiglass tablet with cylindrical holes.
Describes the mathematical relationship between the absorbed dose at the entrance of the detector and the output signal of the detector. In contrast to the S-shaped curve of a film-screen system, most digital detectors show a linear relationship.
A system of creating digital radiographic images that utilises a storage-phosphor plate (instead of film) in a cassette. Once the plate is exposed, a laser beam scans it to produce the digital data that is translated into an image.
DICOM=Digital Imaging and Communications in Medicine
A set of protocols describing how radiology images are identified and formatted. DICOM is manufacturer-independent and was developed by the American College of Radiology and the National Electronic Manufacturers' Association. The standard emphasises point-to-point connection of digital medical imaging devices. DICOM 3.0 is the current version. http://medical.nema.org/
A European Research Consortium on: Measures for optimising radiological information and dose in digital imaging and interventional radiology. See also.http://dbs.cordis.lu/fep/FP5/FP5_PROJl_search.html
A general term referring to radiation dose (i.e. either radiation dose to a patient or to an imaging device). The word `dose' is used in this document whenever the context is not specific to a particular dosimetric quantity. When the context is specific, the name or symbol for the quantity is used [e.g. absorbed dose at a point (D), mean absorbed dose in an organ or tissue (DT), dose-area product (DAP), air kerma (Ka), etc.]. See Appendix B for more complete definitions of the dosimetric quantities used in this document.
Dynamic range (of an x-ray detector)
The ratio of the maximum and minimum dose that can be accepted by an imaging device without deteriorating or distorting the image. Sometimes this dynamic range is expressed as the ratio of the maximum and minimum pixel values (only valid for a linear relationship between absorbed dose at the entrance of the detector and pixel value). Flat-panel detectors have a dynamic range of 104 (from 1 to 10,000), and film-screen systems have a dynamic range of approximately 101.5 (from 1 to 30) (Hennigs et al., 2001).
The ratio of the maximum and minimum absorbed doses from the projected radiation field at the entrance plane of the x-ray detector. The dynamic range of an image depends mainly on the properties of the anatomical region, but can also be influenced by technique factors (e.g. beam quality).
A physical approach for a general description of imaging performance that compares the signal-to-noise ratio (SNR) in the image to the SNR in the radiation field. A DQE value of unity describes an ideal detector. A DQE value of 50% means that the detector of interest requires twice the absorbed dose to generate an image with the same SNR fidelity as the ideal detector. The DQE is usually given as a function of spatial frequency. The DQE of a real detector also depends on beam quality and absorbed dose level.
Healthcare record stored in electronic format.
A general term referring to the act of irradiating a patient for a clinical purpose, or irradiating a device that converts radiation exposure into useful clinical images.
Term used in relation to the absorbed dose to the phosphor plate. After reading the image with the laser system, the signal histogram is computed, and the dose level or exposure index is determined from the pixel values using a logarithmic relationship (Huda et al., 1997).
Gradation curve (see Characteristic curve)
Graininess
Related to noise perception. The impression of inhomogeneity of an apparently (or expected) homogeneous area due to noise sources (quantum noise, electronic noise, fixed structures, stochastic anatomical fluctuations, etc.).
Shades of perceived brightness values that are generated by the display device. Each pixel value of the digital image is mapped to a brightness value. For optimum mapping of pixel values into perceived grey shades, the conversion should utilise the concept of the `grey-scale standard display function' as described in the DICOM standard.
A system used to store and retrieve patient information. The HIS is an integrated computer-based system that may include or be linked to laboratory and radiology information systems (denoted as LIS and RIS, respectively).
Represents the whole set of pixels forming an image. The matrix size is the number of rows and columns in a digital image (e.g. 
). The matrix size of the image cannot be identical to the matrix size of the x-ray detector.
JPEG is a standard for data compression. It offers data compression of between two and 100 times. http://www.highend3d.com/rosetta/?cat=j
A parameter used in image processing algorithms to define the impact of the neighbouring pixels on the pixel of interest. The kernel can be interpreted as a calculation scheme usually written in a matrix form, where the pixel currently being processed is in the centre (e.g. a kernel of 3
×3 specifies that only the information of the adjacent eight pixels is used to recalculate the content of the centre pixel).
A network of a small number of computers whose reach is limited (usually within a building or campus). The computers are linked to allow access and sharing of data and computer resources by users.
Feature on fluoroscopic systems allowing the continuous display of the last image acquired following termination of any exposure period.
A method of compressing and storing a digital image in such a fashion that the original image can be completely reconstructed without any data loss.
A generic term used to describe an imaging device such as magnetic resonance imaging, computed tomography, ultrasound, digital radiography, computed radiography, mammography, etc.
Describes how well the contrast of an object with defined size is preserved in the image. A MTF value of unity means that the contrast has not been degraded at all. A MTF value of zero means that the detail is no longer present in the image. The MTF is usually given as a function of spatial frequency (see ICRU, 1986).
The minimum frequency that will faithfully sample an analogue signal. This is always twice the maximum frequency of the signal to be sampled. In practice, significantly higher sampling frequencies are used in order to stay well above the Nyquist frequency and therefore avoid the chance of producing unwanted artefacts.
Original image
The image obtained (read-out signal of flat-panel detector or storage-phosphor system) after all device-specific corrections.
A system capable of acquiring, transmitting, storing, retrieving, and displaying digital images and relevant patient data from various imaging sources, and capable of communicating the information over a network.
A digital image is formed by single pixels. Each pixel represents the image signal at a defined location in the image. As images are usually rectangular in shape, the pixel location is represented by a column and a row co-ordinate.
The physical structure in an x-ray detector consisting of all means to convert the local x-ray exposure into an electrical signal is also referred to as a `pixel'. Depending on the detector technology, these pixels may have different shapes and sizes, or represent a laser spot size as in storage-phosphor systems. The sensitive area forming the image signal is important in the conversion to an electrical signal, and this area is referred to as a pixel aperture.
The distance between the centres of the pixel apertures. The pixel spacing determines the spatial resolution limit of the device.
A numerical code assigned to a pixel. It represents, on an arbitrary scale, the absorbed dose in the x-ray detector.
Basically the conversion of the analogue, continuous quantity into a discrete (digital) numerical code. The difference between the discrete and the (real) analogue value is called the quantisation error and limits the achievable low-contrast resolution. This quantisation error should always be significantly lower than the local signal variation due to quantum noise.
A grouping of standard disk drives together with a controller to create storage that acts as one disk to provide performance beyond that available from individual drives.
Read-out signal of flat-panel detector or storage-phosphor system. The term `raw data' is often used to emphasise that they are unprocessed.
A system that supports the information processing and business requirements of radiology departments and freestanding image centres.
The physical process of assigning a single value to a pixel by integrating the signal captured by the pixel aperture. Due to the finite size of the pixel aperture, details that are smaller than the pixel aperture are blurred.
The visual impression of the quality of reproduction of small details.
Scaling of the image signal with regard to the absorbed dose to the detector, by performing the digital processing in a way that the brightness of the displayed image is kept on a defined constant level, independent of the exposure.
The ratio of noise-to-picture signal information.
An image processing function for enhancement (or suppression) of anatomical structures depending on the spatial frequency (i.e. the size) of the structure. The most common application is the enhancement of small details by increasing the sharpness impression (so-called edge enhancement). The function is usually controlled by two parameters:
(i) the kernel size in pixels (for controlling the size for enhancement); and(ii) the amount (e.g. as a percentage) (for controlling the enhancement).
Other applications of spatial filters are dynamic range compression or noise suppression (smoothing).
Any signal can be composed of a series of harmonic (sine and cosine) waves. An image can be interpreted as a composition of an infinite number of periodic sine and cosine waves. A short wavelength (equivalent to high spatial frequency) corresponds with small detail, whereas a long wavelength (equivalent to low spatial frequency) corresponds with large objects in the image. The relationship between spatial frequency and detail size is inversely proportional. In order to avoid confusion with the term `time frequency', `spatial frequency' is used. A common unit is line pairs per millimetre (lp/mm).
Following the sampling theorem (Beutel et al., 2000), at least two sampling periods are necessary to detect the frequency of a signal properly. This requires that the sampling frequency be at least twice that of the highest spatial frequency component in the x-ray image (see Nyquist condition). The sampling frequency of the device is determined by the pixel spacing.
The relative `dose' required to obtain images with appropriate optical density (i.e. blackening). The higher the speed class, the lower the `dose' required to obtain the given optical density (i.e. the image).
The ability of an image intensifier to produce a bright image. With digital systems, gain is one of the factors that need to be optimised in order to produce optimal image quality with an acceptable radiation dose.
Allows customisation of the dimensions of the x-ray beam using numerical methods without the need to expose a patient or a phantom. This option is available for some fluoroscopy digital systems.
Due to technical limitations, many display devices are not able to show the full image matrix generated by the x-ray detector. Image viewing software supplies zooming algorithms to handle this situation. It should be noted that dropping pixels when fitting the image to the display area can cause aliasing artefacts, and these should be suppressed by image processing (controlled spatial filtering to limit the frequency content of the displayed submatrix). Most workstations provide a specific `full-resolution' mode that displays exactly one image pixel by one display pixel without any interpolation (`1:1 scaling').
1. Introduction
Main points
1.1. Purpose of this document
(1) This document is addressed to radiologists, radiographers, medical physicists, and all other interested individuals who are familiar with conventional film-based radiography and fluoroscopy. No specific physics, electronic, or computer knowledge is necessary to understand the basic principles of digital radiography and fluoroscopy or their clinical applications. This document discusses the advantages and disadvantages of digital imaging, aspects of radiation dose, elements of image quality, equipment costs, and new regulations and standards. Other digital techniques, i.e. computed tomography (CT) and digital subtraction angiography (DSA), are beyond the scope of this document.
1.2. Introduction
(2) In the last decade, the replacement of conventional fluoroscopic and radiographic equipment with digital imaging systems has increased rapidly in developed countries. The conversion from conventional to digital image acquisition necessitates additional equipment. Processing, reading, storage, and the distribution of digital images requires computers, digital networks, and laser imagers or picture archiving and communication systems (PACS). The wide dynamic range of digital detectors avoids the requirement for repeat examinations due to over- or underexposed images. However, these dynamic ranges, and post processing, also make it more difficult to recognise over- or underexposure as the images may be of diagnostic quality in either case. The dynamic range of flat-panel detectors can achieve a dose reduction of 30–50% without any decrease in diagnostic performance. Also, the dynamic range provides an opportunity to tailor the patient dose and image quality for a specific medical imaging task, e.g. in paediatric examinations. Additionally, digital techniques allow remote diagnosis and expert consultation by teleradiology.
(3) Digital radiology cannot be recommended for all countries. A minimum level of electronic and mechanical engineering expertise and infrastructure is necessary to support such systems. Film-based conventional radiology, with wet image processing, is simple, stable, and, at present, less expensive.
1.3. History
(4) For more than 100 years, x rays have been used with radiographic films for medical imaging. New developments in computer applications in the last decade have fostered new techniques that have extended the spectrum of radiological imaging significantly. Despite the rapid progress of radiology in the area of magnetic resonance imaging (MRI) and CT, large-projection or plain-film radiography from lung, skeletal, and gastrointestinal examinations still represent up to 80% of all examinations, and are the major part of the daily routine in radiology.
(5) Following the introduction of DSA in combination with fluoroscopic systems at the beginning of the 1980s, digital image intensifiers were introduced into the field of projection radiography. This new method was used primarily for investigations of the gastrointestinal tract and other fluoroscopic examinations. In the following years, the obvious advantages (high image quality, potential for lower patient dose, direct availability of images, simple handling, and the other advantages of digital image information format) led to an almost complete replacement of the film-screen technique in image-intensifier-based systems.
(6) In the mid-1980s, storage-phosphor radiography was introduced as a new digital technique for wall- or table-mounted Bucky systems and bedside examinations. The initial significant technical challenges, financial costs, and limited image quality delayed the wide clinical application of this technique until the beginning of the 1990s. Today, storage-phosphor-based systems play a fundamental role in the field of digital projection radiography.
(7) More recently, flat-panel detectors have been developed that enable direct digital registration of image information at the detector. Since 2000, these detectors have become available commercially and they are slowly being introduced into the clinical routine. Dynamic flat-panel detectors are already replacing image intensifiers in routine clinical practice.
(8) An expanding field of research and clinical application is occurring in digital mammography. The spatial resolution of these systems needs to be higher than that for all other digital systems. The initial flat-panel detectors or storage-phosphor systems developed for mammography produce a spatial resolution of 5–8 line pairs per mm. This resolution is still 50% less than that for conventional film-screen systems.
(9) Although spatial resolution is an important consideration in the evaluation of the image quality of the acquired data, the impact of noise needs attention and is of particular importance when considering noise-limited, low-contrast imaging tasks, e.g. detection of microcalcifications and masses in mammography. Complete evaluation of the image quality of the acquired data requires an appreciation of the imaging task, and the measurement of some fundamental parameters such as grey-scale transfer, resolution, noise, and incident radiation exposure.
1.4. What is digital radiology?
(10) Conventional radiographic images are based on the exposure and processing of films. Hence spatial resolution and grey levels are analogue values. The dynamic range or latitude of film is limited by the maximum optical density that the film can produce. Digital radiography and fluoroscopy, which do not have this limitation, utilise a matrix of discrete numerical values to represent an image.
(11) A matrix is a square or rectangular area divided into rows and columns like a chessboard. The smallest element of a matrix is called a pixel (from picture element). The location of each pixel in a matrix is encoded by its row and column number (x,y). Each pixel of a matrix is used to store a binary number with a range of 8–16 bits (bit is short for binary digit). Eight bits allow the storage of integer values between 0 and 255; 16 bits allow values between 0 and 65,535. Small grey-level ranges of 8 bits are used for ultrasound or film printing with laser imagers, and 10–14 bits are used for digital fluoroscopy or digital radiography. These numerical values are related to the individual grey levels of an image. The contrast in a digital image is represented by the difference in the numerical pixel values in different areas of the image.
1.5. Differences between intrinsic digital modalities and digital radiology
(12) Digital radiography and fluoroscopy are new techniques that are capable of replacing film-based image acquisition. In contrast, intrinsic digital modalities such as CT, MRI, DSA, single photon emission CT (SPECT), and positron emission tomography (PET) have no preceding analogue acquisition. Computers were necessary to obtain digital images from x-ray projection data or radio frequency signals. Conversion of conventional films into digital images by means of digitisers is pseudo digital radiology.
1.6. Digitised conventional films
(13) Conventional films may be converted into a digital image format and stored electronically by use of a digitiser. Modern digitisers provide sufficient spatial and grey-level resolution for nearly lossless conversion of information. However, the use of digitisers for conversion with digital storage of all images in a radiology department cannot be recommended. High-quality digitising and the creation of a correct digital patient and examination header is time consuming and adds additional expense. Hence, digitising films can only be recommended in selected cases such as the integration of previous digital examinations into a digital archive (i.e. PACS). Other reasons may include the saving of conventional film copies (if a patient leaves the hospital for further external treatment), and digitising of mammograms for post processing. The use of film digitisers has proven useful in mammography. In this particular application, the digitised data serve as a significant adjunct to the physician.
(14) Digitising of films in non-digital departments for creation of teaching files is an important application, as is digitiser-based or video-camera-based image acquisition for teleradiology. Several commercial attempts to introduce film-based radiographic x-ray systems with integrated digitisers have not been successful.
1.7. Plain radiography
(15) Cassette-based film-screen systems have proved to be a stable technique. The advantages are good image quality, a simple acquisition technique, and moderate costs. Disadvantages are a small dynamic range, the lack of post processing, exposure requirements, and the availability of radiographic film at only one location.
(16) The major difference in all digital radiography systems is the technical approach to imaging, i.e. how the physical process of x-ray detection is realised. The following techniques are available:
(17) Storage-phosphor was the first available technique for digital radiography. The storage plates are exposed in cassettes with standard dimensions. Hence, no change of generator, x-ray tube, and wall- or table-mounted Bucky system is necessary. Even bedside examinations and other special projections are possible.
(18) In 1994, the selenium-cylinder technique was the first introduction of a solid-state detector into chest radiography. It has a high efficiency for x-ray detection that is comparable with modern flat-panel detectors. The application is limited to examinations with a vertical patient position (chest, abdomen, and spine).
(19) CCD systems represent a very small share of the market. The image of a luminescent screen is recorded with CCD cameras or devices and converted into digital images.
(20) The latest developments in digital radiography are flat-panel detectors with direct or indirect conversion of x rays into electrical signals. These detectors provide the highest quantum efficiency, excellent image quality, and enable a substantial patient dose reduction [with the appropriate quality assurance (QA) programme]. Portable flat-panel digital radiography units have been evaluated for neonatal imaging; substantial dose reduction, requiring one-quarter of the patient dose compared with conventional radiography, was found, with rapid availability of images for timely clinical decision making (Samei et al., 2003).
1.8. Digital fluoroscopy
(21) Both digital and conventional fluoroscopic systems are based on the use of image intensifiers (except new flat-panel systems). In conventional systems, the output screen of the image intensifier is projected by an optical lens onto a film. In digital systems, the output screen is projected onto a video camera system or a CCD camera. In addition to the digital acquisition, some systems provide additional cassette holders for exposure of conventional films or storage-phosphor plates. Patient dose can be reduced by digital functions, e.g. last image hold (LIH), virtual collimation, pulsed fluoroscopy, image processing, etc.
1.9. Comparison of digital and non-digital techniques
(22) In comparison with conventional radiology, digital radiology needs new components such as a digital detector, computers, networks, laser imagers, workstations, and digital archives. When introducing digital radiography in a radiology department, some aspects of the workflow will need to be changed. Another main difference is the relationship between exposure and presentation of an image. In conventional film-based radiography, there is a well-known relationship between the film exposure and the result of black, grey, or white images. Both under- and overexposures result in changes with reduced contrast. This can be simplified by: ‘an excessive exposure is black, an inadequate exposure is white’. In digital radiology, the grey levels of an image are normalised and the brightness of an image is independent of the mean absorbed dose in the imaging device. Normal levels of absorbed dose produce good or normal images, and substantial reductions in absorbed dose increase the pixel noise of an image. Further advantages of digital radiology are the features of brightness adjustment with ‘window/level’, magnification, post processing, and various types of digital communication. The critical issue is an overexposure that may not be noticed by the radiologist or the radiographer because the images are still of good quality.
1.10. Image quality and exposure levels in digital radiology
(23) Many studies have compared exposure levels and image quality of digital and conventional radiography. They have concluded that the diagnostic information from digital radiography is superior to conventional radiography if the exposure parameters used are equivalent (or for equivalent image quality, digital radiography requires less exposure). All digital techniques are characterised by a much higher dynamic range, preventing over- or underexposure to a large extent. The spatial resolution of all digital systems is about half that of conventional film-screen systems. Despite this, good diagnostic image quality can be secured by postprocessing methods.
1.11. Further aspects of digital radiology
(24) In the last decade, the number of digital radiology systems installed has increased substantially worldwide, replacing many conventional film-based systems, especially in hospitals. Only digital radiology allows the change from wet chemistry to dry laser imagers.
(25) The use of digital radiology equipment requires special training of radiologists, radiographers, and medical physicists. Despite the advantages of digital systems, there are additional risks of undetected patient pathology and occurrence of specific digital artefacts. Some of these artefacts may be misinterpreted as pathology in a patient.
(26) The physical principles of digital image generation and processing require new approaches, such as the use of specific ‘digital phantoms’ and derived parameters such as ‘exposure index’. The postprocessing algorithms of different manufacturers should be harmonised. Access to the original image is necessary to control the physical detector quality.
(27) The decision to select a specific technique depends on the clinical environment. For small hospitals or institutes with low or moderate examination frequencies, storage-phosphor systems are well-suited, flexible, multipurpose tools. Flat-panel detectors are more expensive, so they are more appropriate if there are large numbers of chest examinations or Bucky-table-based examinations.
(28) Industrial advances have already led to cheaper flat-panel detectors with image acquisition rates that can be used for fluoroscopy, and there is the prospect of new storage-phosphor detectors with an aligned crystal structure and, hence, higher quantum efficiency.
(29) Digital radiology is, at present, more expensive than film-based radiology. These higher costs are justified if all of the following advantages of the new digital technique are considered and incorporated into the daily routine: superior image quality or less patient dose; increased speed and improved workflow; teleradiology and expert consultation; reduction of archive space; and shorter patient examination times with increased comfort.
PII: S0146-6453(04)00002-8
doi:10.1016/j.icrp.2004.02.001
© 2004 ICRP. Published by Elsevier Inc. All rights reserved.
