Radiation Safety & Testing in CT

Keeping radiation dosages in check is a priority for CT scans of adults and especially pediatric patients who run more risks of radiation exposure because they are still growing. Check out the variety of techniques, notably on souped-up multislice scanners, to measure and control dose while producing excellent images.

The use of computed tomography is on the rise. In February 2003, the United States Food and Drug Administration (FDA) reported that the use of computed tomography (CT) for diagnostic studies, cancer staging, treatment planning and real-time visualization during intervention had risen dramatically based on data collected for the Nationwide Evaluation of X-ray Trends (NEXT) Survey. Twenty-nine percent of all CT units in the U.S. are equipped to perform multislice spiral scanning, according to the survey.

David J. Brenner, PhD., DSc, professor of radiation oncology and public health at Columbia University in New York compares the statistic of close to 58 million CTs completed in 2000 with the four million CT scans accomplished in 1980. Brenner continues that although the NEXT report did not provide an exact number of CT studies in pediatrics, it's on the order of 10 percent, or roughly six million scans.


From the perspective of radiation dose safety, several concerns arise.

The broad issue involves the fact that CT examinations deliver a higher radiation dose to the patient than a conventional radiograph because, by its design, CT produces several radiographs from different directions that are then combined to produce the image. While radiation dose to adults is of major significance, the implications for pediatric patients assume critical proportions.

If the CT scanner is set to deliver the normal adult milliamp per second dose (mAs) to a child, the resulting dose will be effectively higher for the child because an adult body provides more shielding, explains Brenner. "For the same mAs, the organ dose can be three times higher, depending on the size of the child."

Brenner describes two converging factors that make increased dose more dangerous to children from a carcinogenic risk perspective. Radiation renders its greatest deleterious effect on dividing cells. Children and neonates, as part of the normal growth process, produce large numbers of rapidly dividing cells. Therefore, children are more sensitive to radiation than adults. Given the fact that children are thinner than adults, often the quality of the image can be maintained even with reduced radiation dose.

Brenner notes that one way to reduce dose dramatically for children is to use other imaging modalities in place of CT. He says research studies estimate that approximately one third of the CT scans performed can be replaced by other imaging exams or alternative diagnostic strategies. For example, in the diagnosis of appendicitis, ultrasound scans or just observing the child over several hours can will provide the diagnostic information required. Many children who present with marked abdominal tenderness become automatic candidates for a CT scan. While these studies do provide excellent diagnostic information, the benefit may not outweigh the risk from increased radiation exposure.

The major CT vendors have recognized the risks involved, and have developed methods to address these concerns for both adults and children.

Philips Medical Systems approached these issues from several directions. Intellibeam technology places two titanium filters and a wedge-shaped filter into the port of the x-ray tube of the Brilliance scanner to eliminate low energy x-ray skin dose that does not contribute to the quality of the image, according to Hugh Morgan, PhD, research scientist in the CT engineering department at Philips. "Typically, our system will deliver about 40 percent less skin dose than the competition, but will not reduce image quality," says Morgan.

Jim Green, Philips' senior vice president and general manager of CT business, says that their multislice scanners help to reduce dose through increased coverage.

Philips has just introduced the Brilliance 40, a 40-slice version that can provide 25 mm of coverage at 0.625 slice thickness, and 44 mm of coverage if the slice thickness is set at 1.2 mm. The system is currently being used for coronary artery analysis and stroke assessment as well as other studies. They anticipate that it will significantly reduce dose for pediatric patients because since the scan is accomplished in a few hundred milliseconds, it removes the motion artifact that often serves as a primary reason for the need to re-scan a child. Coupling the speed of the exam with the fact that at its 44mm setting, it could image an entire organ in a child, the dose reduction becomes significant.

Specifically to address pediatric concerns, Philips developed special infant and small child phantoms used in dose and image quality testing of all of their CT scanners. They adopted pediatric protocols designed to deliver adequate diagnostic dose without delivering higher radiation dosages to these little patients.

Siemens Medical Solutions has incorporated pediatric scan protocols with mAs settings adapted to the patient's age and weight on all of their scanners, according to Andre Hartung, head of CT Product and Clinical Marketing Management. "For example, for newborns we use only 15 percent of the dose of the respective adult protocol. For all contrast studies, we have established the use of lower tube voltages to reduce the amount of dose used without impairing the diagnostic image quality." Additionally, their current family of scanners includes an automatic exposure control that selects the optimal exposure for every slice of the patient and every viewing angle that allows significant dose reduction in adult and especially in pediatric exams.

Bryan Westerman, PhD, clinical sciences manager for Toshiba America Medical Systems, describes the automatic exposure control on their Aquilion 16. "We give the radiologists a range of image qualities and once they decide what is appropriate for that particular examination, the system takes over and automatically inputs all of the parameters that are appropriate to keep the dose down, while still maintaining the image quality," Westerman says. These activities are accomplished with a "scout" scan (TAMS calls it a scannergram) that measures the attenuation of each individual patient. That initial scan provides all of the input data to set the machine's systems.

The Aquilion generates 0.5 mm slices, which improves detail in image quality. Westerman notes that just six years ago, all of the scanners produced 5 to 7mm slices, so multislice CT by design had helped to improve image quality while reducing the number of repeat scans, thereby reducing dose.

In 2002, GE Medical Systems (now GE Healthcare) expanded its Color Coding for Kids CT program, that was originally designed to improve the medical care of children in the emergency room, to include this program on all of their LightSpeed and high speed Qx/i CTscanners. Medical professionals assign the child to one of eight color zones based on weight or length. Information for each color zone indicates size-specific dosages.


Stanley H. Stern, PhD, health physicist in the Radiation Programs Branch of the Center for Devices and Radiological Health for the FDA explains that the rapid progress in the technologic advances in CT scanning has presented problems for establishing standards for dosage management. His group is collaborating with the International Electrotechnical Commission (IEC) to develop voluntary standards for modern CT equipment.

One of the first concepts to be addressed by the FDA and IEC is the Computed Tomography Dose Index (CTDI) that all manufacturers follow to determine the dose volume for any given protocol. The CTDI is an index of system radiation output and absorption in acrylic reference material (known as a phantom).

Stewart C. Bushong, ScD, professor of radiologic science at the Baylor College of Medicine in Houston asserts, that every CT system should be evaluated on a routine basis for image quality and patient dose by a board-certified medical physicist. Bushong describes the testing process where a medical physicist uses a special phantom to establish the dose both on the periphery and in the center to create the CT dose index weighted (CTDI-w). This assessment evaluates dose distribution throughout the patient during a CT exam. Having assessed the patient dose, then the medical physicist will position the ACS test object for evaluation of several image quality factors, such as spatial and contrast resolution and performance specifications such as linearity and uniformity that have been identified by the American College of Radiology Physics Commission

Linearity has to do with the ability of the CT system to track different types of tissue. While this is not critical for image assessment in most diagnostic scans, it is absolutely essential if this CT imaging system is to be used for quantitative CT such as bone mineral studies or cardiac scoring for calcified plaque in the coronary artery.

Uniformity deals with the ability of the CT system to show the same value of the CT number throughout the slice. For example, if water is zero, the number should remain zero throughout the entire slice thickness and throughout that slice wherever water is imaged.

Toshiba's Westerman explains that all of their scanners are tested in the factory, with those test results shipped out with the system. They are re-checked on a regular basis during their preventive maintenance activities to insure proper functioning of the machine. Dose may not be measured every time, but a medical physicist hired by the facility will establish a routine for dose testing.

Siemens' Hartung describes their Acceptance Testing, which includes image quality, dose and other safety issues after installation. Then the customer conducts daily or monthly constancy testing procedures on the scanners and Siemens service professionals. All of these procedures are incorporated in the system software, which runs automatically, evaluates all of the data and gives warnings if anything is out of tolerance.

In addition, Hartung describes that the Siemens scanners have multiple built-in safety features that continuously control the scan parameters and the quality of measured data. Tube parameter settings and the radiation output are monitored for every scan and stored in a file, which is automatically evaluated on-site, and is returned to the factory with the respective tube after a tube exchange.


While IT professionals may not have specific responsibility related to radiation safety testing per se, as members of the radiology team, they need to know about testing activities. They may also be called upon to make decisions about what specific portions of the files should be archived.

The primary concern for IT professionals as related to CT scans involves the management of huge data sets generated by this equipment. Philips' Jim Green relates that with cardiac imaging and coronary artery analysis, there could be volume scans with 3-4000 axial slices in them. Each slice is a 512 X 512 image that is between 0.3 and 0.5 megabytes per slice. Multiplied by 3K images, the file size would approach 1.5 gigabytes for a single study on one patient. If only a few representative slices and a 3D rendered image that provides the three-dimensional view of the area of interest were archived, the size of the file would be reduced to about 5 megabytes of data.

State law, or the institution's risk managers, may mandate decisions about which images should be retained. Green suggests that IT personnel are key decision-makers for these issues.

In terms of the integration of their scanners, Green says, "We designed the system and verified that we can communicate with all of the top PACS and primary HIS/RIS systems. Usually the IT person knows how to configure the system, but at installation time, we help them set it up so that it runs properly in the environment of the hospital."

Toshiba's Westerman echoes the concerns about managing the size of files. Comparing a chest exam, he says that with a single slice scanner, a chest exam would generate about 20 slices, or 10 megabytes of data. Complete that same study on a multislice scanner, between 300 and 400 images would be generated, and if they reconstruct those images with a different reconstruction algorithm, that number must be doubled again to 300 megabytes. He describes run-off exams where they scan from neck to feet, and generate 1,200 images or 600-megabyte files.

Columbia University's Brenner suggests that it would be extremely helpful if IT professionals could begin to collect data about actual dose administered during CT scans for future studies about the carcinogenic risks of these tests.

"The risks we talk about these days are extrapolated from situational results from A-bomb survivors. The doses are not that dissimilar, because there are large numbers of survivors who received relatively low doses that are similar to doses from CT," explains Brenner. "But it would be nice someday to be able to do some epidemiologic studies directly from people who have had CT scans. If IT could keep that information somehow, it would facilitate studies in the future."

Brenner describes a potential study that would look at the development of leukemia in a population of patients who have had CT scans, compared to a control group who had not been scanned. In general, it takes as much as 20 to 25 years for a cancer to develop, but with leukemia, the time span is only two or three years.


Radiation testing is a critically important activity to reduce potentially hazardous radiation doses from CT. Routine testing by board-certified medical physicists is the best method of accomplishing these vital tasks.