HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Willis, C. E. Articles by Slovis, T. L. Search for Related Content PubMed Articles by Willis, C. E. Articles by Slovis, T. L.

The ALARA Concept in Pediatric CR and DR: Dose Reduction in Pediatric Radiographic Exams—a White Paper Conference Executive Summary¹

¹ From the Texas Children’s Hospital, Edward B Singleton Diagnostic Imaging Services, 6621 Fannin (MC 2-2521), Houston, TX 77030-2399 (C.E.W.); and Wayne State University, Detroit Medical Center, Children’s Hospital, 3901 Beaubien Boulevard, Detroit, MI 48201-2196 (T.L.S.). Received October 25, 2004; accepted October 26. Address correspondence to C.E.W. (e-mail: cewillis@texaschildrenshospital.org) or T.L.S. (e-mail: tslovis@med.wayne.edu).

Editor’s Note: In view of the importance of the topic, we are publishing, with permission, this editorial that appeared in Pediatric Radiology 2004; 34(suppl 3): S162–S164.

ALARA, an acronym for as low as reasonably achievable, is a philosophy of radiation-dose management. As the Society for Pediatric Radiology (SPR) continues to promote ALARA in pediatric computed tomography, we have begun to examine other imaging modalities where the potential for dose reduction exists. Computed radiography (CR) and digital radiography (DR) are acquisition systems that replace conventional screen-film (SF) systems for projection radiography. In conventional radiography, the amount of radiation needed to produce an acceptable image is specific to the SF system and chemical processing conditions. In CR and DR, the acquisition process is separated from the display process, allowing these systems to produce acceptable images over a wide range of exposures. Unfortunately, the disconnection of acquisition from display also introduces the potential for systematic overexposure.

In order to examine the risk that CR and DR pose in pediatric radiography and the potential for dose reduction, the SPR organized a second ALARA Conference. The conference was held in Houston, Texas, February 28, 2004 and was attended by 77 pediatric radiologists, medical physicists, radiologic technologists, imaging scientists, and engineers. The conference was made possible by unrestricted grants from six CR and DR manufacturers. The faculty consisted of four academic speakers, five industry speakers, and a speaker from the Food and Drug Administration. The industry speakers were provided a list of questions in order to focus their lectures on specific features of their products that were relevant to dose management in pediatric radiography.

Findings
Research indicates an increased risk of childhood acute lymphocytic leukemia from plain film studies and an increased risk of fatal breast cancer from scoliosis series. The linear, no-threshold model, which states that no level of radiation exposure is without consequence, is currently the best estimate of risk. The younger the patient at the time of exposure, the greater the risk of developing a fatal cancer. Children are roughly an order of magnitude more sensitive to radiation than middle-aged adults.

Overexposure in CR and DR is quite common. There are a number of contributing factors. The disconnection of acquisition from display in CR and DR allows these systems to compensate for variations in exposure factor by automatically rescaling the image to provide a relatively consistent appearance. Underexposed images typically have a grainy, mottled appearance that causes radiologists to reject the images. Over-exposed images, on the other hand, have a crisp, sharp appearance that is favored by radiologists. Faced with the prospect of repeating images, radiologic technologists tend toward overexposing. This phenomenon, so-called "exposure factor creep," has been documented in the CR literature.

The feedback mechanism of density that radiologic technologists are so familiar with is arbitrary and meaningless in CR and DR. Any digital image can be rescaled into any desired density range. Almost every CR and DR system provides a derived numerical indicator of radiation dose. Unfortunately, there is no standardization of the mathematical form or units of these indicators. The indicators depend on proper calibration of the acquisition systems, are subject to interferences, and require interpretation to translate into meaningful feedback for radiologic technologists. Some systems actually allow the technologist to modify the indicator before releasing the image, limiting the utility of the indicator as a record of dose delivered. In some cases, the values of these indicators are not reported on the images, and are often not passed along with the digital images through the electronic image distribution system, confounding oversight efforts by the radiologist. There is evidence in the literature that when exposure indicators for CR are incorporated into a quality control program, patient doses are moderated.

A dichotomy exists in manufacturers’ approaches to dose feedback, roughly divided between detectors that are independent of x-ray generating equipment and those that are integral to x-ray generating equipment. Because of difficulties in getting exposure technique data from an undefined array of x-ray equipment, independent detectors tend to calculate a dose indicator based on the amount of radiation reaching the detector. Integrated detectors tend to calculate a dose area product (DAP) or kerma area product (KAP) value from the exposure conditions. DAP and KAP are independent of source-to-image distance (SID), but unfortunately are expressed in units unfamiliar to American practitioners. There are three exceptions to this categorization: one manufacturer of an independent detector and two manufacturers of integrated units have no exposure indicator, and one manufacturer of an integrated detector calculates detector exposure in addition to KAP.

Manufacturers are reticent about assigning a speed class to their CR and DR devices. These devices can certainly produce images at a wide range of exposure levels, so they can be considered variable speed systems. Density is embedded in the definition of radiographic speed: If density is meaningless in CR and DR, so is speed. However, every radiation detector system is designed with some expectation of a radiation exposure level. This expected exposure level is the basis of construction of any radiographic technique guide, or phototimer setup. Without explicit guidance from the manufacturer, and without interpretable feedback from the acquisition system, CR and DR practitioners are free to develop exposure guidelines that violate the ALARA principle.

Anticipated benefits from medical imaging procedures are tightly coupled with image quality, which is based upon image fidelity and image intelligibility. Image fidelity depends on subject contrast, image resolution, image noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR). In CR and DR, processing of the raw digital image is absolutely necessary in order to render an intelligible image. The extent of image processing that can be applied depends on the SNR and CNR. The radiation dose necessary to achieve a given SNR and CNR is determined by the efficiency of the radiation detector, which is best expressed in the detective quantum efficiency (DQE).

The DQE of CR and DR detectors can be compared to SF detectors. Some CR and DR systems have DQE inferior to FS, while others have equivalent or superior DQE, meaning that they should be able to produce similar image fidelity with equal or less radiation dose. Manufacturers have been working on CR systems with improved DQE.

Appropriate image processing is crucial in producing the optimal pediatric CR or DR image. Unfortunately, there is little standardization of the nomenclature or methods of image processing offered by CR and DR manufacturers. The amount of image processing that can be applied to an image is specific to its SNR and CNR. Processing parameters typically constitute a large number of obscure variables with adjustable values that allow a much larger numerical range than is clinically useful. Practitioners are faced with the choice of either accepting the default processing settings supplied by the manufacturer, established under unknown exposure conditions to produce an appearance to satisfy unknown observers, or to take on the arduous trial-and-error task of customizing processing for their local conditions.

Most CR and DR manufacturers recognize that pediatric patients are unique and have developed special provisions for pediatric exams, including image processing and accommodations for special exams like scoliosis.

Training programs for physicians, medical physicists, and especially radiologic technologists have lagged behind the clinical implementation of CR and DR. This means that the front-line personnel responsible for confronting the issue of dose and image quality are ill equipped to address the problem. The lack of standardization in nomenclature complicates communication of technical information between manufacturers and practitioners.

Recommendations
1. The conference faculty was unanimous in recommending a team approach to dose management in CR and DR. The team should include the active participation of a radiologist, medical physicist, radiologic technologist, biomedical engineer, manufacturer service engineer, manufacturer applications engineer, and manufacturer imaging scientist.

2. Training of radiologic technologists in CR and DR technology and practice was identified as a specific weakness. Aggressive combined efforts are needed between the manufacturers and radiologic technologist professional organizations and training programs to improve the level of awareness among those with hands on both patients and equipment.

3. Discipline is needed in nomenclature for CR and DR technology. Density is still used by some manufacturers to indicate the output grayscale value of their detectors. Speed class is used for gain settings of some systems. Sensitivity and S are used as the name of one exposure indicator. These terms have definitions based in screen-film technology.

4. Dose feedback is absolutely necessary for dose management in CR and DR. Indicators of dose received by the detector are needed as well as DAP or KAP, in order to independently distinguish patient dose and image-quality issues. Dose indicators must be protected from modification by the operator, must be passed to the electronic distribution system in the DICOM header, and picture archiving and communication systems (PACS) must be able to read and display these indicators to radiologists. Dose indicators should be standardized in order to be interpretable in real time. A proposal to initiate standardization of CR and DR exposure indicators has been submitted to the Science Council of the American Association of Physicists in Medicine (AAPM).

5. The nomenclature of digital image processing also needs to be standardized. The position of some manufacturers, that their methods of image processing are proprietary is unacceptable to practitioners. CR manufacturers have made great strides in recent years in providing technical references on the mathematical basis of their processing. DR manufacturers have been less forthcoming. Image processing is intimately associated with exposure conditions: Changing either without knowledge of consequences constitutes bad practice.

6. Each manufacturer needs to provide practitioners with a number that represents the radiation dose expected at the detector in order to produce an adequate image. From that number, practitioners can work backward to develop technique guides, and can make informed decisions about whether to operate at either a lower or higher dose to the detector.

7. Standardization can improve the use of ionizing radiation. Standardization of mammography led to a decrease in mean glandular dose from 14 to 1.8 mGy with concurrent improvement in image quality. The process of standards development begins with observations, reports, and publications, such as this one. At this point, we may enter a cycle of discussion with increased awareness and education. It is possible to continue with individual responsibility for addressing the problem. Voluntary standards may be promulgated by professional organizations, such as the SPR or the AAPM. In the case of mammography, voluntary standards from the American College of Radiology were translated into mandatory standards by the U.S. government in the form of the Mammography Quality and Standards Act (MQSA).

This article has been cited by other articles:

				H. Datz, A. Ben-Shlomo, D. Bader, S. Sadetzki, A. Juster-Reicher, K. Marks, T. Smolkin, S. Zangen, and M. Margaliot The additional dose to radiosensitive organs caused by using under-collimated X-ray beams in neonatal intensive care radiography Radiat Prot Dosimetry, July 1, 2008; 130(4): 518 - 524. [Abstract] [Full Text] [PDF]

This Article Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Willis, C. E. Articles by Slovis, T. L. Search for Related Content PubMed Articles by Willis, C. E. Articles by Slovis, T. L.