Medical Physics and Informatics Original Research

Transcription

1 Medical Physics and Informatics Original Research Jaffe et al. Radiation Dose for Body CT Medical Physics and Informatics Original Research Tracy A. Jaffe 1 Terry T. Yoshizumi Greta Toncheva Colin Anderson-Evans Carolyn Lowry Chad M. Miller Rendon C. Nelson Carl E. Ravin Jaffe TA, Yoshizumi TT, Toncheva G, et al. Keywords: abdominal imaging, CT dose, MDCT, pulmonary angiography DOI: /AJR Received December 31, 2008; accepted after revision March 10, R. C. Nelson is a consultant for GE Healthcare, Inc. 1 All authors: Department of Radiology, Duke University Medical Center, Erwin Rd., Box 3808, Durham, NC Address correspondence to: T. A. Jaffe (jaffe002@mc.duke.edu). AJR 2009; 193: X/09/ American Roentgen Ray Society Radiation Dose for Body CT Protocols: Variability of Scanners at One Institution OBJECTIVE. The objective of our study was to determine, using an anthropomorphic phantom, whether patients are subject to variable radiation doses based on scanner assignment for common body CT studies. MATERIALS AND METHODS. Twenty metal oxide semiconductor field effect transistor dosimeters were placed in a medium-sized anthropomorphic phantom of a man. Pulmonary embolism and chest, abdomen, and pelvis protocols were used to scan the phantom three times with GE Healthcare scanners in four configurations and one Siemens Healthcare scanner. Organ doses were averaged, and effective doses were calculated with weighting factors. RESULTS. The mean effective doses for the pulmonary embolism protocol ranged from 9.9 to 18.5 msv and for the chest, abdomen, and pelvis protocol from 6.7 to 18.5 msv. For the pulmonary embolism protocol, the mean effective dose from the Siemens Healthcare 64- MDCT scanner was significantly lower than that from the 16- and GE Healthcare scanners (p < 0.001). The mean effective dose from the GE 4-MDCT scanner was significantly lower than that for the GE 16-MDCT scanner (p < 0.001) but not the GE scanner (p = 0.02). For the chest, abdomen, and pelvis protocol, all mean effective doses from the GE scanners were significantly different from one another (p < 0.001), the lowest mean effective dose being found with use of a single-detector CT scanner and the highest with a 4-MDCT scanner. For the chest, abdomen, and pelvis protocols, the difference between the mean effective doses from the GE Healthcare and Siemens Healthcare scanners was not statistically significant (p = 0.89). CONCLUSION. According to phantom data, patients are subject to different radiation exposures for similar body CT protocols depending on scanner assignment. In general, doses are lowest with use of scanners. S ince the mid 1990s, the number of CT procedures performed each year in the United States has increased more than 10%, but population growth has been less than 1% per year [1]. As the use of CT has increased in both routine and emergency care [1 9], patient exposure to ionizing radiation has increased dramatically. Although it represents only 15% of total radiologic procedures, CT accounts for almost 50% of the total effective dose (ED) from all diagnostic radiologic studies [8, 10]. CT of the chest, abdomen, and pelvis accounts for approximately two thirds of the collective radiation dose [1]. In addition, a considerable percentage of both inpatients and outpatients undergo repeated imaging [3, 4], exposing individual patients to increased radiation burdens. Given the increase in both the incidence and frequency of CT, it is important to appreciate the factors affecting radiation dose. Doses can be expected to differ on the basis of protocol design and the type of scanner used. Continued technologic advances in CT, particularly progress in MDCT, have stimulated efforts to optimize imaging protocols [11 17]. Until recently, these efforts have been largely concentrated on maximizing image quality, and less attention has been paid to radiation dose. Many radiology departments have a variety of scanners at their disposal, ranging from those designed more than a decade ago to high-end MDCT scanners capable of stateof-the-art cardiac imaging. In spite of the brisk technologic advances in CT technology, including mechanisms to reduce radiation dose, older scanners have not been rendered AJR:193, October

2 Jaffe et al. obsolete. Efficient patient throughput depends on flexibility in scanner assignment, and although some technically challenging angiographic studies require a 16- or scanner, at some institutions, CT pulmonary arteriography and imaging of the chest, abdomen, and pelvis may occasionally be relegated to older scanners. Do patients assigned to different CT scanners on the basis of throughput decisions such as scheduling and scanner availability receive the same radiation doses regardless of scanner designation? To our knowledge, there has been no study of variability in radiation dose during routine body CT among multiple scanners in one radiology department. The purpose of our study was to explore differences in ED for a variety of CT scanners operated with routine clinical CT protocols for pulmonary embolism (PE) and CT of the chest, abdomen, and pelvis. Materials and Methods Phantom and Detector Placement A commercially available anthropomorphic phantom designed to simulate a medium-sized man (model 701-D, CIRS) (Fig. 1) was used to measure absorbed organ doses. The phantom was 173 cm in height and weighed 73 kg. At the level of the thorax, the anteroposterior by transverse dimensions of the phantom were cm. Additional breast tissue (height, 5 cm; diameter, 13 cm) from a female phantom made by the same manufacturer as the male phantom was attached to the male phantom for measurement of breast dose. Twenty metal oxide semiconductor field effect transistor (MOSFET) dosimeters (model 1002RD, Best Medical) with an active detector area of µm (total dimensions, 2.5 mm width 1.3 mm thickness 8 mm length) were placed in defined anatomic locations in the chest, abdomen, and pelvis as designated by the phantom manufacturer (Table 1). Each detector was calibrated at a given CT beam energy, and individual calibration factors for all 20 detectors were stored in a database. Detailed calibration methods and validation of MOSFET methods have been described by Yoshizumi et al. [18]. Scanning Parameters The phantom was scanned with two protocols PE and chest, abdomen, and pelvis on a variety of CT scanners, ranging from a singledetector helical scanner to scanners. Scanner profiles and protocols are listed in Table 2. All GE Healthcare scanners had a gantry aperture of 70 cm. The Siemens Healthcare scanner had a gantry aperture of 78 cm. The chest, abdomen, and pelvis protocols of the GE scanners had been optimized for noise and image quality at our institution and the details published [12, 14, 15]. The PE protocols of the GE scanners were those used at our institution and optimized in the literature [16, 19, 20]. The protocols for the Siemens scanner were those recommended by the manufacturer. Modification of the quality reference tube current time setting for the PE protocol was based on our thoracic radiologists requests to decrease noise (increased to 350 s from the recommended 150 s). We do not use our single-detector CT (SDCT) scanner for pulmonary angiography, thus it was not included in the PE portion of this study. Fixed tube current was used on the SDCT scanner, and automated tube current modulation was used on the other scanners. Differences in automated tube current modulation methods were based on scanner age and software availability. On the 4- and 16-MDCT scanners, Auto (GE Healthcare) was used; on the two scanners, either Smart (GE Healthcare) or Care- Dose 4D (Siemens Healthcare) was used. All imaging was performed at 120 kvp. For both the PE and chest, abdomen, and pelvis protocols, the phantom was imaged on each scanner three times, and an average organ dose was calculated. A Fig. 1 Anthropomorphic phantom of man. A, Photograph shows additional breast tissue. Metal oxide semiconductor field effect transistor detectors are in place. B, CT scanogram shows detectors in place. ED was calculated for each scanner for both protocols on the basis of International Commission on Radiologic Protection 60 weighting factors. To determine image noise, we manually placed a circular region of interest (ROI) (1,465 mm 2 ) in the heart for the PE protocol and within the lower abdomen for the chest, abdomen, and pelvis protocol (Fig. 2). The result was recorded as the SD of attenuation values in the ROI. Statistical Measurements For statistical analysis of dosimetry, one-way analysis of variance was used, and Tukey s studentized range test was applied for multiple comparisons among all of the scanners. Noise values were compared by use of the F test for variances. For each replication, the noise value (ROI SD) was squared and divided by its degrees of freedom to yield a chisquare variable. The three chi-square variables from the three independent replications were summed to a single (composite) chi-square variable with degrees of freedom equal to the sum of the degrees of freedom of the individual chi-square variables. In comparison of two scanning protocols, the ratio of the composite chi-square variables had an B 1142 AJR:193, October 2009

3 Radiation Dose for Body CT TABLE 1: Locations of Organ Dose Metal Oxide Semiconductor Field Effect Transistor (MOSFET) Detectors Dosimeter Organ Slice Location No. Depth (mm) MOSFET No. 1 Skin 20 A1 2 Breast (right) 3 o clock position 12 A2 3 Thyroid A3 4 Thymus A4 5 Lungs (left, middle) (left) 5 A5 6 Bone marrow, ribs (right) 3 B1 7 Bone marrow, thoracic and lumbar spine F distribution under the null hypothesis of equal noise. The F distribution was used to compute p for comparison of the noise values of the different protocols. We did not compare noise values at the individual replication level because the large number of such multiple comparisons would require application of a highly conservative level of significance to each comparison and would have severely limited the power of each comparison. A statistical software package (SAS 9.1, SAS Institute) was used for all statistical comparisons, and p < 0.05 was considered to signify a statistically significant difference. Results The ED results for all scans are shown in Figure 3. The mean ED for the PE protocol ranged from 9.9 to 18.5 msv and for the chest, abdomen, and pelvis protocol from 6.7 to 18.5 msv. For the PE protocol, the mean ED from the Siemens Healthcare 64- MDCT scanner (scanner 5) was significantly lower than that from the GE Healthcare 16- and scanners (scanners 3 and 4) (p < 0.001). The mean ED from the 4-MDCT scanner (scanner 2) was significantly lower B2 8 Bone marrow, lumbar spine B3 9 Bone surface B4 10 Lungs, lower (left) 5 B5 11 Esophagus C1 12 Adrenal glands C2 13 Liver C3 14 Stomach C4 15 Spleen C5 16 Kidney D1 17 Intestine (upper left) D2 18 Ascending colon D3 19 Bone marrow, pelvis D4 20 Urinary bladder D5 than that from the 16-MDCT scanner (scanner 3) (p < 0.001) but not from the GE 64- MDCT scanner (scanner 4) (p = 0.02). For the chest, abdomen, and pelvis protocol, all mean EDs calculated from the GE scanners (scanners 1 4) were significantly different from one another (p < 0.001), the lowest mean ED being found on the SDCT scanner (scanner 1) and the highest on the 4-MDCT scanner (scanner 2). The difference in mean ED from the GE and Siemens scanners with the chest, abdomen, and pelvis protocol was not statistically significant (p = 0.89). Noise values for both the PE and chest, abdomen, and pelvis protocols are shown in Figure 4. For the PE protocol scans, the SD of noise ranged from 8.7 to 15.6 HU; for chest, abdomen, and pelvis scans, the SD of noise ranged from 9.7 to 15.8 HU. In comparisons of the Siemens Healthcare and GE Healthcare scanners, noise was significantly greater on the Siemens scanner with the PE protocol and significantly lower with the chest, abdomen, and pelvis protocol (p < 0.05). Discussion Appropriately, the radiology community is becoming increasingly concerned with the radiation dose attributable to CT. The objective of our study was to determine whether patients receive significantly different radiation doses based on random scanner assignment for common indications for body CT. The protocols used in our study had been either optimized at our institution [12 15, 19] or been widely cited in the literature [16, 20 25]. Most of these protocols were designed to produce high image quality with less attention paid to resultant radiation dose. For this study, we further modified the protocols in an effort to reduce dose. Most radiation dose reduction efforts have been focused on alterations in tube current, namely use of automated tube current modulation. Wide variations in patient radiation absorption occur with changes in CT projection angle and anatomic region, and the projection with the most noise initially determines the amount of noise on the final image [26]. It is possible to reduce the dose for other projections without increasing noise on the final image. Automated tube current modulation addresses this dose reduction with variation in tube current along a specified axis; the x- and y-axes are angular, and the z-axis is longitudinal. The methods and applications of automated tube current modulation have been well explored, the documented dose reductions in adults ranging from 20% to 60% [26 32]. With the exception of imaging with the SDCT scanner, for which tube current modulation was not available, we used automated tube current modulation for all imaging in this study. Differences in patient doses at CT can be linked to multiple causes. Imaging parameters set by operators vary, including number of acquisitions performed, exposure technique, and slice thickness [8]. Two scan parameters that play a role in determining dose slice thickness and table speed were not held constant in this study, and it is important to note the effects on dose. For example, in the change from the SDCT scanner to the 4-MDCT scanner, section thickness decreased from 7 mm to 5 mm, which would increase noise if not for a compensatory increase in tube current and, consequently, radiation dose. Some of this dose increase can be offset by the automated tube current modulation mode as, in the case of Auto (GE Healthcare), a constant noise index is set. Use of tube current modulation alters our AJR:193, October

4 Jaffe et al. TABLE 2: Scanners and Protocols Current Time Setting (s) Rotation (s) Table Speed (mm/ rotation) Exposure Time (s) Pitch Quality Reference Effective Minimum and Maximum Current () Noise Index Current () Voltage (kv) Section Thickness (mm) Reconstruction Interval (mm) No. of Images Obtained Detector Configuration Scanner Protocol 20.2, :1 CAP 27, 22, , 210, 170 1, CT/I (GE Healthcare) , :1 15 PE Auto 2, QX/I (GE Healthcare) , , :1 15 CAP , Auto , : PE Auto 3, LightSpeed 16 (GE Healthcare) , , : CAP , Auto , :1 55 PE Smart 4, VCT (GE Healthcare) , 700 5, :1 55 CAP , Smart PE , Definition 64 (Siemens Healthcare) a CAP , , , Note CAP = chest, abdomen, and pelvis; PE = pulmonary embolism. a Dual-source scanner used in single-source mode for this study. ability to directly compare the 4-MDCT scanner dose with that of the SDCT scanner, on which automated tube current modulation is not an option. Overbeaming (extension of the x-ray beam beyond the detector array) also may add to the higher dose on the 4-MDCT scanner. All of these factors likely influenced the increase in dose from the SDCT to the 4-MDCT scanner. With the increase from 4-MDCT to 16-MDCT, there is a decrease in section thickness for the chest, abdomen, and pelvis protocol from 2.5 mm to mm. This difference is the major contributor to the increase in dose; however, this phenomenon is largely offset by a faster table speed for the 16-MDCT scanner (15 vs 17.5 mm/rotation) for the chest, abdomen, and pelvis protocol. Although we still observed a dose increase, the difference was not statistically significant. For the PE protocol, the statistically significant increase in dose with the 16-MDCT scanner compared with the 4-MDCT scanner was mostly attributable to a decrease in table speed for the 16-MDCT scanner. The increase in table speed accounts for the decreased dose for the chest, abdomen, and pelvis and PE protocols of the 64- MDCT scanners compared with the 16-MDCT scanners because all other imaging parameters, including section thickness, were unchanged. Inherent differences in the scanners themselves, including design, manufacturer, and proprietary method of automated tube current modulation, can contribute to the range of doses for CT [33]. Scanner design has been found to influence dose in multiple ways. For example, scanner geometry can be used to decrease dose, which was manifested in our study with lower doses for the SDCT scanner in the chest, abdomen, and pelvis protocol. First, the x-ray penumbra used in MDCT shows greater scatter contribution than does that used in SDCT [34]. Second, radiation dose is proportional to the inverse of the squared distance between the radiation source and the point of measurement, or tubeto-isocenter distance. The tube-to-isocenter distance for the SDCT scanner (scanner 1) was 630 mm, longer than that in all of the MDCT scanners used in our study (e.g., GE Healthcare MDCT scanners, 541 mm, or 16.5% closer; Somatom Definition, Siemens Healthcare, 595 mm, or 5% closer). When all other scanning parameters are maintained, the radiation dose from an MDCT scanner is higher than that of an SDCT scanner owing to the shorter tube-to-isocenter distance. In comparison of the Siemens Healthcare and GE Healthcare scanners, the mean ED for the PE protocol was significantly lower for the Siemens scanner. This difference is thought to result from the differences in vendor-specific style of automated tube current modulation. GE uses an operator-selected image noise setting (noise index) and a look-up table to match the patient s specific attenuation values measured on a scout image to the tube current values for each gantry rotation. This system is designed to maintain image noise as attenuation values change from one rotation to 1144 AJR:193, October 2009

5 Radiation Dose for Body CT Fig. 2 Anthropomorphic phantom of man. Regions of interest (ROIs) placed manually in soft tissue in same location for each scan. SD of ROI is used to measure noise. A, Transverse CT scan through chest shows ROI in heart. B, Transverse CT scan through abdomen shows ROI in lower abdomen and pelvis. Effective Dose (msv) , GE SDCT PE CAP 2, GE 4-MDCT 3, GE 16-MDCT Scanner A 4, GE 5, Siemens Fig. 3 Graph shows differences in effective dose for pulmonary embolism (PE) and chest, abdomen, and pelvis (CAP) protocols. GE = GE Healthcare, SDCT = single detector CT, Siemens = Siemens Healthcare. Noise (HU) the next. Siemens Healthcare uses a setting called quality reference s to establish image quality level. The operator selects an effective tube current time product typically used for a sample patient with a weight of approximately 80 kg. The noise target is varied on the basis of patient size, and image noise is not kept constant but is adjusted on the basis of an observed impression of image quality. Topograms are used to predict a tube current curve to yield the desired image quality. An online system is used to fine-tune current values during scanning to match patient-specific attenuation at all angles. Because the two manufacturers automated tube current modulation modes are not comparable, it is understandable that there would be inherent differences in doses derived from the two scanners. Because the imaging parameters for the Siemens Healthcare PE protocol suggest that there should be an increase in dose, the major cause of the decreased dose observed with use of this scanner compared with its GE Healthcare counterpart is related to the different current modulation methods, which appear to be more substantial in scanning of the chest than of the abdomen and pelvis. This phenomenon was observed in a multivendor study of radiation dose during MDCT of the chest [35]. It is important to note that our study settings for quality reference s for the Siemens Healthcare PE protocol were higher than those found in the literature or those suggested by the vendor. If we had chosen the vendor-recommended tube current time settings, an even greater discrepancy in dose between the two vendors could have resulted. A final cause of dose variability in CT is patient size. We controlled for this factor by using a single phantom. The larger the patient, however, the greater is the need for increasing energy flux. This increase in patient weight can be overcome with an increase in tube potential [36]. Alternatively, if beam energy is held constant, increases in tube current will be required to overcome the effects of increased body fat. Schindera and colleagues , GE SDCT PE CAP B 2, GE 4-MDCT 3, GE 16-MDCT Scanner 4, GE 5, Siemens Fig. 4 Graph shows differences in noise for pulmonary embolism (PE) and chest, abdomen, and pelvis (CAP) protocols. GE = GE Healthcare, SDCT = single detector CT, Siemens = Siemens Healthcare. [37] found that oversized patients undergoing abdominal MDCT with tube current modulation receive significantly higher doses, especially to the skin, likely because of the greater attenuation of the traversing photon flux through the wider girth of an obese patient. As Schindera et al. noted, if one assumes breast exposure is similar to skin exposure, oversized women are at risk of high radiation doses to the radiosensitive breast tissue, and modification of automated tube current modulation protocols is warranted. There were several limitations to our study. First, the study was performed on a single anthropomorphic phantom of a medium-sized man, so we did not assess doses measured in larger or smaller adult phantoms or those reflecting the pediatric population. It is possible that radiation of more radiosensitive peripherally oriented organs (e.g., breast, thyroid) in women and children may have greater influence on ED than radiation of organs in a man. Second, we were restricted in the number of MOSFET detectors available, so we AJR:193, October

6 Jaffe et al. measured skin exposure with only one detector. Because skin dose varies with xy modulation, our contribution to understanding of the fluctuations in skin dose with each protocol was limited. Third, it was not possible to completely replicate scanning protocols from one scanner to another owing to inherent design differences from scanner to scanner. For each GE Healthcare MDCT scanner, scanning parameters, including detector configuration, slice thickness, and table speed, were modified according to the number of detector rows. When migrating our protocols to the newest Siemens Healthcare scanner, we were unable to replicate the parameters used on its GE counterpart. A final limitation was that although we attempted to optimize body protocols to maximize image quality and radiation dose reduction, further efforts on the part of radiologists may be warranted. Radiologists have become accustomed to demanding the highest image quality. Years of research and protocol refinement have led to protocols that maximize image reconstruction for lesion conspicuity and reader preference at the expense of increased dose. Although we maintained image quality with reduction in tube energy in this study, we did find a statistically significant increase in image noise. Further efforts in dose reduction will lead to the surrender of some of our expectations for image quality. Stated simply, to reduce radiation dose, radiologists will have to become tolerant of image noise. We conclude that phantom radiation data suggest that depending on scanner assignment, patients experience significant differences in radiation exposure for similar body CT protocols. In general, doses are lowest for 64- MDCT scanners and are lower for an SDCT scanner than for 4- and 16-MDCT scanners. Significantly lower doses occur with a PE protocol on a Siemens Healthcare scanner than on the GE Healthcare counterpart, largely owing to differences in automated tube current modulation method. References 1. Mettler FA Jr, Thomadsen BR, Bhargavan M, et al. Medical radiation exposure in the U.S. in 2006: preliminary results of NCRP 6-2. Health Phys 2008; 95: Burge AJ, Freeman KD, Klapper PJ, Haramati LB. Increased diagnosis of pulmonary embolism without a corresponding decline in mortality during the CT era. Clin Radiol 2008; 63: Lee SI, Saokar A, Dreyer KJ, Weilburg JB, Thrall JH, Hahn PF. Does radiologist recommendation for follow-up with the same imaging modality contribute substantially to high-cost imaging volume? Radiology 2007; 242: Broder J, Warshauer DM. Increasing utilization of computed tomography in the adult emergency department, Emerg Radiol 2006; 13: Chen JL, Dorfman GS, Li MC, Cronan JJ. Use of computed tomography scanning before and after sitting in an emergency department. Acad Radiol 1996; 3: Donohoo JH, Mayo-Smith WW, Pezzullo JA, Egglin TK. Utilization patterns and diagnostic yield of 3421 consecutive multidetector row computed tomography pulmonary angiograms in a busy emergency department. J Comput Assist Tomogr 2008; 32: Levin DC, Rao VM, Parker L, Frangos AJ, Sunshine JH. Recent trends in utilization rates of noncardiac thoracic imaging: an example of how imaging growth might be controlled. J Am Coll Radiol 2007; 4: Mettler FA Jr, Wiest PW, Locken JA, Kelsey CA. CT scanning: patterns of use and dose. J Radiol Prot 2000; 20: Boone JM, Brunberg JA. Computed tomography use in a tertiary care university hospital. J Am Coll Radiol 2008; 5: Mukundan S Jr, Wang PI, Frush DP, et al. MOS- FET dosimetry for radiation dose assessment of bismuth shielding of the eye in children. AJR 2007; 188: Schoepf UJ, Holzknecht N, Helmberger TK, et al. Subsegmental pulmonary emboli: improved detection with thin-collimation multi-detector row spiral CT. Radiology 2002; 222: Jaffe TA, Nelson RC, Johnson GA, et al. Optimization of multiplanar reformations from isotropic data sets acquired with 16-detector row helical CT scanner. Radiology 2006; 238: Thomton FJ, Paulson EK, Yoshizumi TT, Frush DP, Nelson RC. Single versus multi-detector row CT: comparison of radiation doses and dose profiles. Acad Radiol 2003; 10: Killius JS, Nelson RC. Logistic advantages of four-section helical CT in the abdomen and pelvis. Abdom Imaging 2000; 25: Gupta AK, Nelson RC, Johnson GA, Paulson EK, Delong DM, Yoshizumi TT. Optimization of eight-element multi-detector row helical CT technology for evaluation of the abdomen. Radiology 2003; 227: Patel S, Kazerooni EA, Cascade PN. Pulmonary embolism: optimization of small pulmonary artery visualization at multi-detector row CT. Radiology 2003; 227: Schueller-Weidekamm C, Schaefer-Prokop CM, Weber M, Herold CJ, Prokop M. CT angiography of pulmonary arteries to detect pulmonary embolism: improvement of vascular enhancement with low kilovoltage settings. Radiology 2006; 241: Yoshizumi TT, Goodman PC, Frush DP, et al. Validation of metal oxide semiconductor field effect transistor technology for organ dose assessment during CT: comparison with thermoluminescent dosimetry. AJR 2007; 188: Hurwitz LM, Yoshizumi TT, Goodman PC, et al. Radiation dose savings for adult pulmonary embolus using bismuth breast shields, lower peak kilovoltage, and automatic tube current modulation. AJR 2009; 192: Wittram C. How I do it: CT pulmonary angiography. AJR 2007; 188: Funama Y, Awai K, Hatemura M, et al. Automatic tube current modulation technique for multidetector CT: is it effective with a 64-detector CT? Radiol Phys Technol 2008; 1: Arakawa H, Kohno T, Hiki T, Kaji Y. CT pulmonary angiography and CT venography: factors associated with vessel enhancement. AJR 2007; 189: Revel MP, Petrover D, Hernigou A, Lefort C, Meyer G, Frija G. Diagnosing pulmonary embolism with four-detector row helical CT: prospective evaluation of 216 outpatients and inpatients. Radiology 2005; 234: Patel S, Kazerooni EA. Helical CT for the evaluation of acute pulmonary embolism. AJR 2005; 185: Kritsanepaiboon S, Lee EY, Zurakowski D, Boiselle PM. CT pulmonary angiography evaluation of pulmonary embolism in children. (abstr) In: Syllabus of the 51st Annual Meeting of the Society for Pediatric Radiology. Reston, VA: Society for Pediatric Radiology, McCollough CH, Bruesewitz MR, Kofler JM Jr. CT dose reduction and dose management tools: overview of available options. RadioGraphics 2006; 26: Papadakis AE, Perisinakis K, Damilakis J. Angular on-line tube current modulation in multidetector CT examinations of children and adults: the influence of different scanning parameters on dose reduction. Med Phys 2007; 34: Tack D, De Maertelaer V, Gevenois PA. Dose reduction in multidetector CT using attenuationbased online tube current modulation. AJR 2003; 181: Rizzo S, Kalra M, Schmidt B, et al. Comparison of angular and combined automatic tube current modulation techniques with constant tube current CT of the abdomen and pelvis. AJR 2006; 186: Mulkens TH, Bellinck P, Baeyaert M, et al. Use of 1146 AJR:193, October 2009

7 Radiation Dose for Body CT FOR YOUR INFORMATION an automatic exposure control mechanism for dose optimization in multi-detector row CT examinations: clinical evaluation. Radiology 2005; 237: Kalra MK, Rizzo S, Maher MM, et al. Chest CT performed with z-axis modulation: scanning protocol and radiation dose. Radiology 2005; 237: Kalra MK, Maher MM, Toth TL, et al. Techniques and applications of automatic tube current modulation for CT. Radiology 2004; 233: Koller CJ, Eatough JP, Bettridge A. Variations in radiation dose between the same model of multislice CT scanner at different hospitals. Br J Radiol 2003; 76: Mettler FA Jr, Huda W, Yoshizumi TT, Mahesh M. Effective doses in radiology and diagnostic nuclear medicine: a catalog. Radiology 2008; 248: Hausleiter J, Meyer T, Hermann F, et al. Estimated The reader s attention is directed to another article pertaining to this topic, which appears on page radiation dose associated with cardiac CT angiography. JAMA 2009; 301: Huda W, Scalzetti EM, Levin G. Technique factors and image quality as functions of patient weight at abdominal CT. Radiology 2000; 217: Schindera ST, Nelson RC, Toth TL, et al. Effect of patient size on radiation dose for abdominal MDCT with automatic tube current modulation: phantom study. AJR 2008; 190:344; [web]w100 W105 AJR:193, October