CT plays a central, increasingly influential role in medical imaging, largely due to the combination of faster scanning and improved image quality available with multidetector CT. We recognize the potential for its benefits and are comfortable with these technological advances, but we are less familiar with accountability for the risks. A few years ago, radiologists were challenged to reconcile the benefits with the risks, and much of our efforts over the ensuing period have focused on the important issues of CT radiation dose and risks, and strategies for managing radiation dose.

It is important to clarify one term. I have chosen, in line with a recent summary from the National Council on Radiation Protection and Measurement,1 to use the term management with respect to CT radiation. While reduction is obviously implied and constitutes one goal, particularly given the often indiscriminant doses that both adults and children receive, management also acknowledges the importance of a balance between radiation and image quality.

WHERE WE WERE

CT has been a powerful tool for more than 30 years. Advances with helical and subsequent multidetector technology have offered expanding and diverse opportunities. These and other technical advances paced and occasionally outran our capacity to know how to best use them.

In the midst of all this opportunity, potential, and great promise,2,3 our attention was redirected. Three articles appeared in the American Journal of Roentgenology.4-6 One pointed out that the CT settings for children were not adjusted based on age or on anatomical region. The mean tube current across all age groups, for example, was relatively high for abdomen (207 mA) and surprisingly higher for chest (213 mA) helical CT.4 Up to this time, manufacturers were reluctant to provide pediatric CT protocols, and published guidelines for size-based pediatric CT were virtually nonexistent. Clearly, we were more interested in what CT could do than in what it shouldn't do.

The 2001 series in the AJR generated intense public reaction regarding the potential risk with radiation and CT.7 A call went out for a broad response from many sectors, including referring clinicians, radiologists, and technologists, other scientific personnel involved in CT imaging, industry, and policy-makers.8

WHERE WE ARE

A few salient facts underscore the current status of CT and radiation. First, some 57,000,000 to 65,000,000 CT examinations are performed in the U.S. each year.1 This translates to one CT for every four to five individuals. The increase in use of CT, currently estimated to be growing at 10% to 15% per year,9 predates the availability of the newest multidetector technology with attendant routine applications, including CT screening.10

Each CT examination can result in a dose of upwards of 30 mSv,9 which is more than 10 times background exposure. It is easy to see why CT accounts for the largest medical source of radiation and, after background, the largest single source of exposure to the world population.1 With recognition of these facts and the perspective that we must be accountable for CT radiation dose (e.g., the as low as reasonably achievable-ALARA-principle), two main thrusts for change have emerged: technology and technique.

Current technological developments for radiation dose management include automatic tube current modulation (ATCM), improved x-ray beam utilization, and improved filters. ATCM is based on the fact that a single tube current is not appropriate given the variation in cross-sectional geometry (wider side-to-side than front-to-back), overall thickness, and regional attenuation. To maintain image quality based on noise, or mottle, for example, more tube current will be required through the upper abdomen than the upper chest. The tube current can thus be modified in the x, y-axis (cross-sectional differences) or the z-axis (regional differences). Minimum and maximum tube current values can be set to ensure ceilings on dose and noise. The type of modulation used and the technical considerations underlying this modulation will vary depending on manufacturer, but the basic goal is the same: reducing the radiation dose without compromising image quality. Recent investigations report a dose savings of as much as 20% to 60% with ATCM.11-14

It is important to understand, however, that ATCM modulation does not necessarily optimize the examination. There is still some operator input. The amount of tube current, as well as minimum and maximum settings, may need to be chosen or adjusted. ATCM simply refines this chosen level. Using ATCM with a level of 300 mA in a five-year-old patient undergoing an abdomen CT, for example, is still in excess of recommendations.15 In addition, ATCM may not be as effective when the patient is very small or is off-center in the gantry (personal communication, M. Kalra, Boston).

Other technical advances include more efficient use of the x-ray beam. MDCT beam geometry is such that some photons are not used for image formation. Work is under way to capture and convert some of this unused beam.16 Another potential advance is the use of noise reduction filter technology. Kalra et al recently reported on this technique, which segments the projection (or raw) data and processes the separated components, recombining the processed data.17 One effect was the reduction in image noise compared with the unfiltered image data. Our preliminary work with this type of filter technology suggests that pediatric body CT might be performed at tube current levels that result in a relatively noisy data set that could be filtered to decrease the noise as if the examination were obtained at a higher tube current (Figure 1). Whether this affect on noise will translate to increase conspicuity has yet to be shown.

Current technique, or procedural, developments include size- or weight-based CT scanning guidelines, as well as preliminary investigations into adjustments in scan techniques based on indication, low dose thresholds for lesion detection, and adjustments in peak kilovoltage (kVp). Pediatric body and brain guidelines have been provided by manufacturers over the past two to three years.14,18-20 Much of this information came in response to adverse publicity about CT radiation dose and cancer risks. One weight-based system, categorized by color groups, was compared with a more traditional weight-based system and found to be significantly easier to use, more error-free (closer adherence to appropriate pediatric settings), and preferred by technologists.21 This system has subsequently been incorporated into routine 16-slice scanning pediatric body protocols.18

Modifications in parameters such as tube current based on scan indication are also becoming available. These modifications are based on the principle that scan quality can be measured, in part, by image noise. The noisier the image, the lower the spatial resolution. In some indications, such as assessment for renal stones, higher image noise is acceptable for diagnosis. A higher noise setting (lower mA) might be used for the set point for ATCM in this setting, as opposed to evaluation for potential small, low-contrast lesions such as hepatic microabscesses or metastases.

Many guidelines for adjustments in scan settings, usually tube current, have been empiric (Figure 2). The core question is really how low you can go with tube current and still provide acceptable diagnostic quality. This is obviously a complex issue and involves a great many variables, including characteristics of the lesion or disorder, size of patient, intrinsic contrast of the region examined, and other settings such as pitch and kilovoltage, as well as local or regional practice standards and personal experience.

One new research tool being applied to answer some of these questions is tube current simulation. With this technology, noise is added to projection data to create an examination that is identical in appearance to one actually performed at a tube current as low as 40 mA.22 This eliminates many of the methodologic difficulties with lower dose CT evaluation, especially serial scans in a single patient. This tool can be applied to virtually any type of CT scan (Figure 3). Results in pediatric body CT to date have shown that reductions of up to 67% in dose can be achieved for certain indications, such as evaluation for large or high-contrast abnormalities.22 A compelling application of this simulation technology is in establishing the levels for ATCM in both adults and children based on size and scan indication.

Peak kilovoltage is another parameter for CT that has been relatively neglected. In a recent survey of the members of the Society for Pediatric Radiology's use of helical CT techniques, Hollingsworth et al reported a kVp of < 110 was used by only 3% of pediatric radiologists for chest CT and 1% for abdomen CT.23 Kilovoltage has an exponential relationship to dose, so that reductions have a greater impact than adjustments in tube current, which are linearly related to dose. A reduction from 140 to 120 kVp, for example, will decrease dose by over 30%. Some empiric evidence in children suggests that 80 or 100 kVp is acceptable in small children and CT angiography.15 Further investigation of kVp, radiation dose reduction, and image quality will likely be forthcoming.

What has been the impact of this information to date? There has been an increase in literature dealing with this issue in radiology journals.1,24 National conferences have dealt specifically with CT, radiation dose, and radiation management.1,25,26 Educational material has also been designed for and sent to referring clinicians.27 CT accreditation by the American College of Radiology is in progress, and this includes radiation dose limits.28 It remains to be shown, however, what the global effect has been on adjustment of CT protocols. The survey of pediatric radiologists cited above was performed just before the issue of CT radiation came to light. Results included the fact that 11% to 26% of body CT examinations of children less than nine years old were performed using a tube current of more than 150 mA. Notably, up to 25% of respondents did not know specific parameters.23 While it is likely that these numbers would be different today, this has yet to be confirmed. We are now conducting a follow-up survey to assess the current state of practice.

WHERE WE NEED TO BE

While progress has been made, a great deal of work remains to be done in CT radiation management. Areas that warrant further efforts include better assessment and documentation of CT radiation dose, continued equipment improvements, additional investigations into refining dose versus image quality, and outcome assessment.

One of the difficulties with determining the radiation dose from CT is defining a standard set of measures. For example, CT dose index (CTDI), an increasingly familiar feature on scanner consoles, can be defined in a variety of ways. Confusion with terms has prompted a call to minimize and standardize measures and terminology.1 In addition, the dose delivered by CT needs to be determined more accurately. Some methods of determining the effective dose equivalent (EDE) in mSv, for example, use as few as two different-diameter acrylic phantoms. The EDE from these phantoms usually does not represent the actual dose to the patient. We recently compared the estimated EDE based on the CTDI and dose length product (DLP) on the console with the actual EDE, which was measured by thermoluminescent detectors in an infant phantom. We found that, for a range of body protocols, the dose was overestimated by up to five times compared with the EDE determined from organ dose measurements. The anthropomorphic phantom is smaller and its absorption of x-rays is different from the cylindrical phantoms used for CTDI and DLP determination, resulting in this overestimate. The danger is that, when the patient is larger, the estimate of EDE will be lower than the actual EDE based on measured organ doses. Preliminary work in developing more appropriate pediatric phantoms for CTDI is addressing this issue.17

It would also be useful to have an automatic annotation of estimates of dose on the scan itself. This could then be archived in the patient's radiation dose record, which could include doses from all imaging examinations and techniques that expose the individual to radiation. It seems inconsistent to have no accounting of CT dose, a large source of radiation.

CT scanner technology will continue to move ahead, with more efficient use of x-rays and better filtering, resulting in improved image quality without a dose penalty. The benefits of increasing detector rows past 16, for example to 128 or greater, would be obvious in terms of reducing some motion artifact and scan duration and increasing coverage. Other benefits such as improved diagnostic information are debatable, but it is worth pointing out that eight- and 16-slice scanning was initially met with some skepticism in terms of these benefits.

As discussed, a recent trend considers scan quality based on image noise. Rather than thinking in terms of the product of tube current and gantry cycle time, or mAs, for an individual patient, the goal is to determine the acceptable amount of noise for an examination. This is a basis of ATCM, but more work in defining the appropriate noise levels for patients of all sizes is needed. We have noted, for example, that noise (standard deviation in Hounsfield units) of 15 can be perfectly acceptable for routine adult image quality, but it is too great for infant body CT, where less than 10 appears to be more appropriate.

Investigations in assessing the cost-benefit ratio are also paramount. Radiation dose management requires additional work in assessing the effect of adjustments in tube current through techniques such as tube current simulation as well as changes in kVp.29 Basically, we still have to determine how low we can go. In the cost-benefit equation for CT, where we now realize radiation is a cost, we must determine the usefulness of CT. Much of the new technology is market-driven,9 and outcome assessment of CT, in terms of health improvement and resource use, needs to be elucidated.

The contemporary practice of CT requires a working knowledge of the costs, especially radiation, and the benefits. While an increasing effort from academic and industry sectors has added to our body of knowledge and advanced technology, there are still substantial gains to be made.

References

1. Linton OW, Mettler FA. National conference on dose reduction in computed tomography, emphasis on pediatrics. AJR 2003;181:321-329.

2. Berland LL, Smith JK. Multidetector-array CT: once again, technology creates new opportunities. Radiology 1998;209:327-329.

3. Rogers LF. Helical CT: the revolution in imaging. AJR 2003;180:883-884.

4. Paterson A, Frush DP, Donnelly LF. Helical CT of the body: are settings adjusted for pediatric patients? AJR 2001;176:297-301.

5. Donnelly LF, Emery KH, Brody AS, et al. Minimizing radiation dose for pediatric body applications of single-detector helical CT. AJR 2001;176:303-306.

6. Brenner DJ, Elliston CD, Hall EJ, et al. Estimated risks of radiation-induced fatal cancer from pediatric CT. AJR 2001;176:289-296.

7. Sternberg S. CT scans in children linked to cancer later. USA Today. January 22, 2001:1.

8. Rogers LF. Radiation exposure in CT. Why so high? AJR 2001;177:277.

9. Frush DP, Applegate K. Computed tomography and radiation: understanding the issues. J Am Coll Radiol. In press, September 2003.

10. Brant-Zawadzki M. CT screening: why do I do it? AJR 2002;179:319-326.

11. Greess H, Nomayr A, Wolf H, et al. Dose reduction in CT examination of children by an attenuation-based on-line modulation of tube current (CARE Dose). Eur Radiol 2002;12:1571-1576.

12. Greess H, Wolf H, Baum U, et al. Dose reduction in computed tomography by attenuation-based on-line modulation of tube current: evaluation of six anatomical regions. Eur Radiol 2000;10:391-394.

13. Tack D, De Maertelear V, Gevenois PA. Dose reduction in multidetector CT using attenuation-based online tube current modulation. AJR 2003;181:331-334.

14. Suess C, Chen X. Dose optimization in pediatric CT: current technology and future innovations. Pediatr Radiol 2002;32:729-734.

15. Donnelly LF, Frush DP. Pediatric multidetector body CT. Radiol Clin North Am 2003;41:637-655.

16. Toth TL. Dose reduction opportunities for CT scanners. Pediatr Radiol 2002;32:261-267.

17. Kalra MK, Maher MM, Sahani DV, et al. Low-dose CT of the abdomen: evaluation of image improvement with use of noise reduction filters-pilot study. Radiology 2003;228:251-256.

18. Morgan HT. Dose reduction for CT pediatric imaging. Pediatr Radiol 2002;32:724-728.

19. Fox SH, Toth T. Dose reduction on GE scanners. Pediatr Radiol 2002;32:718-723.

20. Westerman BR. Radiation dose from Toshiba CT scanners. Pediatr Radiol 2002;32:735-737.

21. Frush DP, Soden B, Frush KS, Lowry C. Improved pediatric multidetector CT using a size-based color-coded format. AJR 2002;178:721-726.

22. Frush DP, Slack CC, Hollingsworth CL, et al. Computer simulated radiation dose reduction for pediatric abdominal multidetector CT. AJR 2002;179:1107-1113.

23. Hollingsworth CL, Frush DP, Cross M, Lucaya J. Helical CT of the body: a survey of techniques used for pediatric patients. AJR 2003;180:401-406.

24. Rogers LF. Low-dose CT: how are we doing? AJR 2003;180:203.

25. Slovis TL. Conference on the ALARA (as low as reasonably achievable) concept in pediatric CT: intelligent dose reduction. Pediatr Radiol 2002;32:217-218.

26. Slovis TL. ALARA conference proceedings. The ALARA concept in pediatric CT-intelligent dose reduction. Pediatr Radiol 2002;32:242-244.

27. Society for Pediatric Radiology and National Cancer Institute. Radiation and pediatric computed tomography: a guide for health care providers. 2002. www.cancer.gov/cancerinfo/causes/radiation-risks-pediatric-CT [last accessed 7/02/03]

28. www.acr.org/dyna/?doc=departments/stand_accred/accreditation/index.html [last accessed 7/02/03].

29. Mayo JR, Aldrich J, Muller NL. Radiation exposure at chest CT: a statement of the Fleischner society. Radiology 2003;228:15-21.

DR. FRUSH is an associate professor of radiology and chief of pediatric radiology at Duke University Medical Center in Durham, NC.