The primary operators of CT scanners are radiologic technologists. Until 2007, specific education regarding the physics of CT equipment was not part of the routine curriculum for radiologic technologists (Morrison G, American Society of Radiologic Technologists, personal communication). Therefore, many technologists operating CT equipment may benefit from additional training on CT unit–specific dose reduction features. As a minimum qualification, radiologic technologists operating CT scanners should be registered by the American Registry of Radiologic Technologists (ARRT). Radiologic technologists should participate in professional development programs, such as the CT Basics course provided by the American Society of Radiologic Technologists. All technologists who perform CT should work toward postprimary CT certification through the ARRT. Technologists should take advantage of the free online technologist education modules available on the Image Gently Website (www.imagegently.org), and should “take the pledge” on the Image Gently Website to show their commitment to patient care and safety value.

The technical aspects associated with proper image quality at reduced radiation doses to the pediatric patient during CT are not trivial and can be confusing to the radiologist and the technologist. Radiologists and technologists of any facility should have access to a qualified medical physicist who ensures that the technical aspects of CT are properly understood and applied to the unique design of the facility's CT scanners and the unique clinical practice of the department [5]. The qualified medical physicist may be on staff at the facility or hired as an outside consultant. At a minimum, the qualified medical physicist should be certified by the American Board of Radiology or the American Board of Medical Physics in a diagnostic imaging physics subspecialty (or in radiologic physics) [6, 7].

Commitment to obtaining accreditation of all of a facility's CT scanners by the American College of Radiology (ACR) promotes quality. Clinical staff members, radiologists, radiologic technologists, and medical physicists all must be properly certified to obtain accreditation. The accreditation process is rigorous; radiologists, medical physicists, and radiologic technologists learn about their equipment and determine whether or not their protocols fall within national guidelines. This quality indicator is reassuring for informed parents and patients who come to your facility. Recently, the ACR approved a special logo in conjunction with the Image Gently campaign that identifies a facility as compliant with pediatric protocols as set forth by the ACR (Fig. 1).

CT is fast and accurate and may be the most appropriate examination for a child. There should be little hesitancy, for example, in performing CT in a child with severe injuries after an automobile accident. Scientific studies have shown that imaging in these situations decreases morbidity and mortality and the incidence of exploratory surgery. In other scenarios, there is strong evidence to support the use of ultrasound or MRI rather than CT [5]. In patients with chronic illness, such as inflammatory bowel disease, strategies such as MR enterography are being explored. Children with possible ventriculoperitoneal shunt malfunction often undergo repeated head CT. As an alternative, Ashley et al. [8] describe the use of rapid, single-sequence MRI as an alternative.

Engage your clinical colleagues and enlist their support to promote radiation protection for children. Provide imaging consultations to the referring physician. This action can help decrease the number of inappropriate studies [9]. Question those indications that do not make sense. A CT requested for an “ovarian mass” in a 12-year-old girl raises the question of whether an alternative imaging test, in this instance ultrasound, would be more appropriate. In fact, in this example, a radiograph showed a tooth, likely within a dermoid tumor. Ultrasound was performed, and the patient proceeded to surgery without the need for additional exposure to ionizing radiation. Giving a conference to clinicians on radiation protection for children can be a first step in educating referring physicians (a free Microsoft PowerPoint presentation is available on the Image Gently Website, www.imagegently.org). The use of publicly available ACR appropriateness criteria (www.acr.org/ac) and evidence-based scientific data provide other dose-saving strategies. For example, a recent large-scale study identified children at very low risk of clinically important brain injury after head trauma [10]. Large-scale studies such as this provide referring doctors with the scientific evidence necessary to order examinations that are indicated for the specific clinical scenario

To develop appropriate pediatric radiation doses, one must verify that “adult-size” technique factors do not deliver estimated radiation doses larger than those recommended by the ACR's CT Accreditation Program [11]. One universal CT technique for the adult patient cannot be used with all vendors' CT equipment. Differences in CT scanner design (e.g., bow-tie filters, focal spot-to-detector distance, detector efficiency, and others) make it impossible to estimate patient radiation dose on the basis of technique factors alone. Consequently, a qualified medical physicist should measure the radiation output from the CT scanner and estimate the radiation dose for an adult-size abdomen and head on the basis of the facility's standard adult radiographic protocols. These estimates become the standard adult abdomen and head radiation doses. Measured dose estimates should be less than the reference dose levels for adults published by the ACR.

Child-size the CT scanning parameters based on the “adult-size” calculations established in step 6. The scanning field of view and collimation should be adjusted to the smaller size of the child. CT parameters should be modified (see step 8) to ensure that the radiation dose to pediatric patients does not exceed the radiation dose given to the baseline adult patient. The universal protocols on the Image Gently Website can be used as a starting point to lower doses. These instructions assume all technique factors (other than tube current or gantry rotation time) remain fixed as techniques are adjusted for pediatric patients. Although several authors have advocated changing other parameters (e.g., peak kilovoltage [kVp]) to reduce dose, these reduction recommendations do not apply if parameters other than tube current–time product (mAs) are changed [12–15]. Work closely with your qualified medical physicist to ensure that image quality is maintained after achieving your desired dose reductions. If you are using the automatic exposure control (AEC) features of your CT scanner, the AEC system may automatically reduce techniques for children, provided the adult baseline is set up properly. This needs to be verified by your qualified medical physicist.

Reducing patient dose in CT must be balanced against changes in image quality. However, increases in the quantum mottle or background noise in your images may still be acceptable. High-contrast images that use IV contrast material during CT angiography may allow increased quantum mottle at reduced radiation doses. As described by Karmazyn et al. [16], a low-dose CT technique may be used to detect renal stones in children without compromising detection. Using a proprietary tool for simulating reduction in tube current, investigators determined that radiation dose could be halved without significantly altering the detection of kidney and ureteral stones.

Because increased quantum mottle affects low-contrast image quality more than high-contrast image quality, dose reductions for low-contrast images may need to be more modest. For example, soft-tissue differentiation (low contrast) requires lower noise in the image and a larger dose. Despite this, Singh et al. [17] further adjusted protocols on the basis of referral indication and number of previous scans. Does the patient have a large abscess from a perforated appendix that could be followed with a limited lower-dose CT when ultrasound cannot be performed? We may see increasingly individualized scanning protocols in the future.

Because general use of CT as a diagnostic tool makes it difficult to adjust parameters to optimize image quality, at a minimum, basic scanning parameters should be adjusted to manage radiation dose to the patient [18, 19].

a. Center the patient in the gantry—Because the entrance dose to the skin of the patient is in part a function of the distance of the skin from the focal spot of the CT scanner (inverse-square law), positioning the patient's body in the middle of the CT gantry reduces the radiation dose to the patient [20].

b. Reduce doses during projection scout (topogram) views—Although the default setting for the scout image acquisition typically may be anteroposterior on CT scanners, the scanner probably allows a posteroanterior projection scout image on the supine patient. This significantly reduces doses to radiosensitive organs, such as male gonads, breast, thyroid, and lens of the eye, located at the exit plane of the patient. Proper adjustment of the high voltage and tube current used for the projection view significantly reduces radiation dose to the patient. O'Daniel et al. [21], measured the radiation exposure from the scout scan on 21 scanners representing 11 different models from three different manufacturers. They determined that by adjusting scanning parameters of the default survey scan from 120 kVp to 80 kVp and changing the tube position from 180° to 0°, radiation exposure could be reduced from all scanners to less than the exposure of a chest radiograph.

c. Axial versus helical mode—Body imaging is typically performed in the helical mode, in which the x-ray beam is continuously on during the scan as the patient anatomy continuously advances through the gantry. This results in the irradiation of a cylindrical volume of patient anatomy; the length of the scan volume equals the scan length, or z-axis. High-end state-of-the-art CT scanners contain programmed collimator blades that attenuate irradiation of tissues that are not imaged adjacent to the ends of the cylindrical volume. If this relatively new collimation feature is not present, tissues at the ends of the irradiated volume are needlessly irradiated. The cylindrical volume of patient data allows reformatting of the images retrospectively from the transverse plane to either the coronal or sagittal plane. 3D models may be reformatted retrospectively. One long scan during helical scanning is better than several regional scans to eliminate scan overlap at the stop and start of adjacent multiregion scans.

Head imaging has typically been performed in the axial mode. The x-ray beam is on for a 360° rotation with the patient stationary. The gantry couch advances the patient's body into the gantry while the x-ray beam is off. The cycle is repeated until the appropriate scan length of the patient anatomy is irradiated. Because the patient couch is stationary during irradiation, programmed collimator blades are not necessary to prevent irradiation of nonimaged patient anatomy. The resolution in the z direction (direction parallel to the long axis of the patient) is not degraded by movement of the patient's body during acquisition. Some manufacturers allow the technologist to control the start of irradiation for each slice. Careful observation of the patient by the technologist allows initiation of each acquisition when the patient is less likely to move.

In pediatric imaging, the advantages and disadvantages of axial and helical imaging must be carefully considered by the technologist, radiologist, and medical physicist. In pediatric imaging, helically acquired head studies or axially acquired body studies may be the correct choice. When the patient is cooperative, helical scanning is typically the mode of choice for body imaging because of the ability to reformat the images to any of the three available planes plus the ability to create 3D models. Because the resolution in the image along the z-axis is not degraded during axial scanning, in some cases, depending on the clinical imaging task, the axial mode of scanning may be preferred.

Some scanners allow initiation of individual images by the operator during axial scanning. This technique may be particularly helpful in the uncooperative patient. If this level of control is not possible, the helical mode may be the best choice because this technique minimizes the time required to collect the entire scan volume.

If the scanner is not designed to spare nonimaged tissues at the ends of the scan volume from irradiation during helical scanning, the axial mode may result in reduction of dose to these organs. Axial scanning with the gantry tilted during head acquisition in some cases may reduce radiation dose to radiosensitive organs, e.g., lens of the eye.

d. Reduce detector size in z direction during acquisition—For both helical and axial scanning, scanning should be performed with the smallest detector element size in the z direction provided by the scanner. If this minimum dimension is 0.5 mm, the scanned voxel of patient tissue is approximately a cube. This allows reformatting of images in the sagittal or coronal planes or in a 3D model without loss of high-contrast resolution relative to the transverse plane. After reformatting, multiple 0.5-mm slices should be combined to increase the volume of the voxel (length) and reduce the quantum mottle in the image without increasing the radiation dose to the patient. Loss of image quality due to partial volume averaging (thick slices) must be balanced against an increase in quantum mottle (thin slices) when selecting the slice thickness at which reformatted images are displayed.

e. Adjust the product of tube current and exposure time—The product of the tube current (rate of x-ray production) and exposure time (duration x-rays are produced) controls the number of x-rays produced during the scan. Changing the mAs directly changes the radiation dose in the same direction and the associated quantum mottle in the images (noise) changes in the opposite direction. The mAs should be adjusted in response to the patient's physical dimensions; larger patients require increased mAs to prevent unacceptable increases in quantum mottle. The required mAs is also dependent on the specific imaging task. When performing high-resolution chest CT, lower mAs (lower dose) can be used to assess airway patency and parenchymal lung disease because high-contrast images are affected primarily by sharpness, not a moderate increase in quantum mottle. Similarly, some have used special low-dose protocols to view the ventricular size and location of the tip of the catheter [22]. On the other hand, higher mAs (higher dose) is required to assess the presence of metastases in the liver, which may be missed on a low-contrast image with increased quantum mottle.

f. When to adjust the kilovoltage—Increasing the kVp increases the energy carried by each photon and results in a more penetrating x-ray beam. A lower kVp decreases patient dose and increases quantum mottle in the image whereas an increase in the kVp has the opposite effect if mAs is unchanged. Typically, the mAs is changed in the opposite direction to the change in high voltage to reduce the degree of change of the radiation dose and quantum mottle in the image [23]. The choice of kVp should be made on the basis of the need for subject contrast enhancement in the image as well as subject size [12–15]. The bony details of the patient's anatomy or soft-tissue studies using an IV or intraluminal contrast agent are increased by a reduction in kVp and increase in mAs to maintain acceptable quantum mottle in the image. Soft tissues of the patient's anatomy imaged without the use of a contrast agent are typically improved by increases in the kVp with appropriate reductions in the mAs to result in reasonable patient doses. To improve bone detail or to perform CT angiography, 100 kVp is reasonable for medium to large pediatric patients. Neonates to small pediatric patients may be imaged at high voltage values as low as 80 kVp; however, 80-kVp images at the maximum tube current of the CT scanner will not produce an adequate number of x-rays to maintain reasonable quantum mottle in the image for larger pediatric patients. To evaluate soft tissues without IV or oral contrast administration, 120 kVp is reasonable for most soft-tissue imaging in children.

g. Increase pitch—Pitch is the ratio of the distance the CT table advances through the scanner during a 360° rotation of the gantry relative to the width of the x-ray fan beam in the z direction. Increased pitch values do not result in reconstruction errors that degrade image quality until a point of anatomy is imaged through less than 180° of rotation. For most scanners, this occurs at pitch values greater than 1.4. The advantage of increased pitch is a reduction in radiation dose if other parameters are not changed because each point of anatomy is irradiated for a shorter time. The radiation dose is proportional to 1 / pitch. By increasing the pitch, the elapsed time from the beginning to the end of data acquisition is reduced. This reduces the chance of motion artifacts and problems with breath-holding. The downside to an increase in pitch is the increase in quantum mottle in the images if other parameters are not changed. The choice of pitch must be balanced with the choice of mAs to result in proper patient dose and image quality [24]. In general, for pediatric body imaging, use a pitch of approximately 1.3–1.4 and a short rotation time (∼ 0.5 second) to minimize total scanning time. Increase the tube current as required to obtain the target patient dose discussed earlier.

h. Manual or automatic exposure control— Most state-of-the-art CT scanners have some level of AEC that is designed to change the tube current (mA) in response to the length of the pathway of the x-rays through the patient's body. Therefore, the mA in the automatic mode changes as the beam rotates between the posteroanterior lateral, anteroposterior lateral, and other projections and as the beam translates along the z direction of the patient's body. The AEC feature is designed to create images with the same quantum mottle regardless of the path length of the radiation through the patient's body [25]. The design of some scanners allows straightforward application of AEC for both adult and pediatric patients. Unfortunately, the design of some CT scanner's AEC is not intuitive and can be difficult for the operator to master for pediatric patients. This automatic mode, when present, can be selected or deselected by the operator. When the automatic mode is off, the tube current operates at constant value regardless of the rotational projection of the beam or the location of the beam along the z-axis of the patient. The AEC mode of the CT scanner should not be used for pediatric imaging if operators do not have verification by their qualified medical physicist via measurement that the use of the AEC mode results in reasonable patient doses. In some instances, use of the AEC mode may increase the patient dose relative to the manual mode.

The radiation risk to the patient is related to both the radiation dose the patient receives and the volume of the patient's body that is irradiated. For this reason, limiting the length of the scout and the scan length of the clinical images is just as important as limiting the radiation dose to the patient. The extent of the scout and the scan should be limited to the area of concern [26, 27]. Overscanning along the z-axis, that is, the length of the patient, increases radiation dose. If scanning of the chest for pectus excavatum is performed, a limited scan length through the area of maximal anterior chest wall depression may be all that is indicated. Moore et al. [28] looked at trauma patients and found little justification for including chest CT in requests for “total body scans.” In the study by Fefferman et al. [29], the authors limit the CT for appendicitis to below the lower pole of the right kidney. In another study, Taylor [30] acquired scans below L3 through the pelvis for appendicitis.

Some scans, such as those assessing hip position after placement of a spica cast may be short in scan length. For chest CT, the scout and scan should cover from the lung apices to the diaphragm. For CT of the abdomen, the scout and scan should cover from the diaphragm to the iliac crests. Inclusion of breast tissue should be minimized. For CT of the pelvis, the scout and scan should extend from the iliac crests to the symphysis pubis and exclude the testes. Specific strategies have evolved for some clinical indications, such as follow-up head CT for ventriculoperitoneal shunt malfunction. Some centers perform these examinations with a shorter scan length by scanning only through the lateral ventricles. Others have used a special low-dose protocol to view the ventricular size and location of the tip of the catheter [22]

Scan only once. Each phase of the CT protocol contributes to the radiation dose. Performing both unenhanced and contrast-enhanced abdominal CT can result in twice the radiation dose for the child [31]. Single-phase scans are usually all that is necessary in children [32]. Unenhanced and contrast-enhanced CT or delayed imaging rarely provide additional information and rarely should be performed except in specific indications.

Have the patient information entered into the CT scanner and the scanning room completely ready before putting the patient on the CT table. This shortens the time the patient is on the table and reduces the chance of patient motion from optimal positioning that may require repeat scanning and radiation exposure. To minimize motion in infants and young children, swaddle the infant. Use distraction tools to improve cooperation and projectors with child-friendly themes, music, toys with flashing lights or music, child-friendly images on the ceiling or walls, singing, counting, and a parent reading and talking to the child through the console all can help reduce anxiety and comfort the child.