Original Research
CT Imaging
February 2008

Effect of Patient Size on Radiation Dose for Abdominal MDCT with Automatic Tube Current Modulation: Phantom Study

Abstract

OBJECTIVE. The purpose of this study was to evaluate in a phantom study the effect of patient size on radiation dose for abdominal MDCT with automatic tube current modulation.
MATERIALS AND METHODS. One or two 4-cm-thick circumferential layers of fatequivalent material were added to the abdomen of an anthropomorphic phantom to simulate patients of three sizes: small (cross-sectional dimensions, 18 × 22 cm), average size (26 × 30 cm), and oversize (34 × 38 cm). Imaging was performed with a 64-MDCT scanner with combined z-axis and xy-axis tube current modulation according to two protocols: protocol A had a noise index of 12.5 H, and protocol B, 15.0 H. Radiation doses to three abdominal organs and the skin were assessed. Image noise also was measured.
RESULTS. Despite increasing patient size, the image noise measured was similar for protocol A (range, 11.7–12.2 H) and protocol B (range, 13.9–14.8 H) (p > 0.05). With the two protocols, in comparison with the dose of the small patient, the abdominal organ doses of the average-sized patient and the oversized patient increased 161.5–190.6%and 426.9–528.1%, respectively (p < 0.001). The skin dose increased as much as 268.6% for the average-sized patient and 816.3% for the oversized patient compared with the small patient (p < 0.001).
CONCLUSION. Oversized patients undergoing abdominal MDCT with tube current modulation receive significantly higher doses than do small patients. The noise index needs to be adjusted to the body habitus to ensure dose efficiency.

Introduction

The effectiveness of automatic tube current modulation to reduce radiation dose and improve image quality has been discerned for abdominal CT [15]. With the technique, tube current is automatically adjusted to the X-ray attenuation of the patient section being scanned to keep the radiation exposure as low as possible and to obtain images with a constant specified image quality [6]. Various types of automatic tube current modulation, including z-axis modulation, xy-axis modulation, and combined z- and xy-axis modulation, have been introduced. State-of-the-art MDCT scanners include xyz-axis tube current modulation, whereby tube current is modulated to patient-specific attenuation in all three planes.
In automatic tube current modulation for abdominal CT, the radiation dose is increased for a larger body habitus [1, 3, 4]. With the advent of high-output X-ray tubes, which are capable of producing peak tube currents up to 800 mA, it is technically feasible to maintain constant image quality over a wide range of patient sizes. Thus in attempts to obtain CT images of consistent quality, the radiation exposure can vary substantially between a thin and a large patient. The potential risk of very high radiation doses to oversized patients when automatic tube current modulation is used has been noted by several authorities [2, 5, 7]. However, the exact extent of the radiation dose received by oversized patients undergoing high-output CT with automatic tube current modulation has not been investigated, to our knowledge. Because obesity is a growing public health problem in many countries and CT plays an important role in the diagnostic evaluation of obese patients with abdominal comorbid conditions, more data are needed to solve this important clinical problem. The purpose of our study was to evaluate in a phantom study the effect of patient size on radiation dose in abdominal MDCT performed with xyz-axis automatic tube current modulation.

Materials and Methods

Anthropomorphic Phantom

An adult-sized whole-body anthropomorphic phantom of a woman (model 702, CIRS) was used to assess abdominal organ doses and image quality (Fig. 1). The phantom was fabricated of epoxy resins for accurate simulation of the physical density and X-ray interaction of various human tissues. The phantom was composed of 38 2.5-cm-thick sectional slabs. The cross-sectional diameter at the level of the upper abdomen measured 18 cm in the anteroposterior direction and 22 cm in the lateral direction.
Fig. 1 Photograph shows female adult anthropomorphic phantom encased with two 4-cm-thick fat rings covering upper abdominal portion. Phantom was placed on wooden board, which was fixed on CT examination table.
To simulate patients of different sizes, the phantom was custom fitted at the level of the upper abdomen (T12–L3) with one or two 4-cm-thick circumferential layers of fat-equivalent material (Fig. 1). Each fat ring had a CT attenuation of –80 H and measured 10 cm in the longitudinal direction (Fig. 2). The fat rings covered four sectional slabs of the upper abdomen. Three types of phantom setups simulated three patients with different body sizes: a small patient (cross-sectional diameter of phantom setup without fat ring, 18 × 22 cm), an average-sized patient (setup with one fat ring, 26 × 30 cm), and an oversized patient (setup with two fat rings, 34 × 38 cm). The cross-sectional dimensions of the three phantom setups were in accordance with published human data on patient sizes [1, 8]. For example, the abdominal anteroposterior by lateral diameter derived from 153 patients in one study averaged 25 × 32 cm [1].

CT

The anthropomorphic phantom with its three setups was scanned on a 64-MDCT scanner (VCT, GE Healthcare). Two scanning protocols (A and B) were used with xyz-axis automatic tube current modulation, which merges z-axis and xy-axis modulation techniques (Smart mA, GE Healthcare). The modulation technique for the z-axis is an automatic tube current modulation feature used to adjust tube current along the z-axis of the patient despite the wide range of attenuation caused by changing patient size, anatomic features, and tissue composition [6]. In contrast, xy-axis modulation is used to adjust the tube current to minimize X-rays over angles that are less important in reducing overall image noise. Thus for anatomic features that can be highly asymmetric, such as the shoulders, X-rays are significantly less attenuated in the anteroposterior direction than in the transverse direction, so the tube current can be substantially reduced in this direction without causing a substantial increase in overall image noise.
Fig. 2 Anteroposterior scout CT image of anthropomorphic phantom shows two 10-cm-wide fat rings covering four sectional slabs of upper abdomen. Fat rings were placed at level of upper abdomen (T12–L3).
A single scout image was used to measure attenuation area and oval ratio, which were used to determine the tube current for the selected CT protocol. For both scanning protocols, the tube voltage was set at 140 kVp, representing the standard tube voltage for routine abdominal CT in our department. For protocol A, a noise index of 12.5 H was chosen, and for protocol B, 15.0 H. The selected noise indexes were recommended by Kalra et al. [1] for routine abdominopelvic CT examinations. The minimum and maximum tube currents were 10 and 715 mA, respectively. At 140 kVp, the maximum tube current allowed by the MDCT scanner was 715 mA. The gantry rotation time for the simulated small patient was 0.5 second, and for the simulated average- and oversized patients, 1.0 second. The longer gantry rotation time was selected to prevent use of tube current settings greater than 715 mA in the two phantom setups with a wider girth. All other CT parameters were kept identical for the two protocols (Table 1).
TABLE 1: CT Parameters
ParameterProtocol AProtocol B
Detector configuration64 × 0.625 mm64 × 0.625 mm
Peak kilovoltage (kVp)140140
Noise index (H)12.515.0
Tube current range (mA)10-71510-715
Gantry rotation time (s)  
    Simulated small patient0.50.5
    Simulated average-sized and oversized patient1.01.0
Beam pitch1.3751.375
Table feed per gantry rotation (mm)5555
Reconstructed slice thickness (mm)
5
5
Longitudinal scan coverage was 15 cm, which included the four sectional slabs of the upper abdomen encased by the 10-cm-wide fat ring plus 2.5 cm cranial and caudal to the ring (Fig. 2). To avoid errors in radiation dose and image quality caused by misplacement of the phantom, great effort was made to use the scanner laser to accurately align the phantom with the gantry isocenter.
One of the authors assessed the mean tube current for the three phantom setups associated with protocols A and B and then multiplied by the gantry rotation time to acquire the mean tube current–time product. The CT dose index was obtained from the operator console of the scanner.

Detector Calibration Method

For measurements of the radiation dose to the upper abdominal organs and to the skin, we used a metal oxide semiconductor field effect transistor (MOSFET) system (model TN-RD-60, Thomson-Nielsen) with high-sensitivity radiology dosimeters (TN-1002RD, Thomson-Nielsen). The MOSFET reader was connected to a notebook computer (Latitude, Dell), and the data were read immediately after each CT exposure.
We calibrated the MOSFET detectors as follows. First, we determined the thickness of copper sheets to achieve the half-value layer (7.24 mm aluminum at 120 kVp) of the CT scanner with a conventional radiographic X-ray tube [9, 10]. Second, we added 0.2-mm copper sheets to the X-ray tube to obtain an equivalent half-value layer of 7.37 mm aluminum at 120 kVp. Third, individual MOSFET detectors were calibrated at our clinical energy level of 140 kVp. During calibration, detectors were placed beside an ion chamber (10x5-6, Radcal). Radiation exposure was read with a radiation monitor (model 9015, Radcal) that had a built-in function for automatic temperature and pressure corrections. Conversion from exposure to absorbed dose was computed by multiplication by an f-factor of 0.94 at 140 kVp [9].

Radiation Dose Assessment and Statistical Analysis

To measure radiation dose to the upper abdominal organs, 13 MOSFET dosimeters were placed in drilled holes in the anthropomorphic phantom at the following three abdominal organ locations: right and left lobes of the liver (five MOSFET detectors), stomach (five detectors), and spleen (three detectors). To avoid scattered radiation from beyond the fat rings, we used only holes that were covered by the fat rings at least 3 cm in the cranial direction and 3 cm in the caudal direction. The three phantom setups with the MOSFET detectors were scanned twice with each of the two protocols. For skin dose measurements, six MOSFET detectors were placed anteriorly, according to the three setups, either on the surface of the anthropomorphic phantom or on the surface of the fat rings. Because the helical beam pattern of MDCT exposes the detectors in a random manner, we took a conservative approach to determining a representative skin dose by accepting the three highest readings of the two CT acquisitions.
The mean radiation dose and its SD were computed for each of the three abdominal organs and the skin. For statistical analysis of the radiation dose, we compared the two protocols within the same phantom setup and different phantom setups within the same protocol. For the latter, we compared the simulated small patient with the simulated average- and oversized patients within protocols A and B. We applied a two-way analysis of variance with the protocol and the phantom setup as the factors. A value of p < 0.05 was considered to indicate a statistically significant difference. All statistical analyses were performed with SAS software system version 9.1.3.
Fig. 3 Diagrams of axial CT images show simulated patients scanned with protocols A and B. Top, small patient (phantom without fat ring); middle, average-sized patient (phantom encased by one fat ring); bottom, oversized patient (phantom encased by two fat rings). As cross-sectional diameter of phantom increases, image quality is maintained in protocols A and B. Left bottom image shows region of interest placement for image noise measurement (circle).

Image Quality Assessment and Statistical Analysis

For image quality assessment, the three phantom setups were scanned once with protocols A and B without the MOSFET detectors to eliminate streak artifacts. Image quality was assessed by an investigator on a separate workstation (Advantage Windows 4.2, GE Healthcare) using reconstructed 5-mm-thick transverse images. For the image quality assessment, image noise was measured by placement of an approximately 2,000-mm2 region of interest (ROI) in the center of the phantom (Fig. 3). Image noise was defined as the SD of the ROI value. Three ROI measurements were obtained on three axial CT images for each of the three phantom setups scanned with the two protocols. Mean and SD were calculated for the image noise values, which were compared among the two protocols and the phantom setups by means of two-way analysis of variance. A difference of p < 0.05 was considered statistically significant.

Results

Within protocols A and B, mean tube current–time product, range of tube current values, and CT dose index all increased in relation to the size of the simulated patient (Table 2). Despite the increasing patient size, the two protocols had similar image noise (p > 0.05) (Table 3). For the three patient sizes, the radiation dose associated with both protocols was always higher for the skin (range, 6.3–78.8 mGy) than for any of the abdominal organs (range, 4.7–44.5 mGy). The radiation doses associated with protocol A, which had a noise index of 12.5 H, were 26.6–34.3% higher than those associated with protocol B, which had a noise index of 15.0 H (Table 4). At the same time, for the three patient sizes, the image noise increased 18.5–21.2% from protocol A to protocol B. This increase paralleled the difference between the 12.5- and 15.0-H noise indexes (20% increase).
TABLE 2: Mean Tube Current–Time Product, Range of the Tube Current Product, and CT Dose Index
CT ProtocolMean Tube Current—Time Product (mAs)Tube Current (mA)CT Dose Index (mGy)
Small patient   
    A102.881-1985.0
    B70.556-1353.5
Average-sized patient   
    A166.7157-17214.6
    B115.2108-1199.7
Oversized patient   
    A576.6468-63147.8
    B
400.1
354-438
33.2
TABLE 3: Radiation Dose and Image Noise
Radiation Dose (mGy)
ProtocolLiverStomachSpleenSkinImage Noise (H)
Small patient     
    A7.4 ± 0.86.4 ± 3.07.4 ± 0.88.6 ± 0.411.9 ± 0.6
    B5.2 ± 1.34.7 ± 1.55.0 ± 0.26.3 ± 0.514.1 ± 0.5
Average-sized patient     
    A20.4 ± 1.218.6 ± 4.219.5 ± 0.831.7 ± 0.811.7 ± 0.7
    B13.6 ± 8.513.0 ± 5.413.5 ± 6.421.3 ± 2.513.9 ± 0.6
Oversized patient     
    A41.1 ± 8.340.2 ± 13.144.5 ± 4.778.8 ± 4.512.2 ± 0.4
    B
27.4 ± 5.4
26.4 ± 9.4
29.4 ± 3.7
54.7 ± 3.4
14.8 ± 0.7
Note—Values are mean ± SD.
TABLE 4: Percentage Difference in Radiation Doses between Protocols A and B and Patient Sizes
Patient SizeProtocolLiverStomachSpleenSkinImage Noise
SmallA vs B-29.4a-26.6a-32.4a-26.7a18.5
AverageA vs B-33.3-30.1-30.8-32.818.8
OversizedA vs B-33.3-34.3-33.9-30.621.2
Small vs averageA175.7190.6163.5268.6-1.7a
 B161.5176.6170.0238.1-1.4a
Small vs oversizedA455.4528.1501.3816.32.5a

B
426.9
461.7
488.0
536.0
5.0a
a
No statistically significant difference (p > 0.05); all other comparisons show a statistically significant difference (p < 0.05).
Within both protocols, the lowest radiation doses occurred in the small patient and the highest doses in the oversized patient (Table 3). Within protocols A and B, the average-sized patient received 161.5–190.6% and the oversized patient received 426.9–528.1% higher abdominal organ doses than did the small patient (p < 0.001) (Table 4). The skin doses increased up to 268.6% for the average-sized patient and up to 816.3% for the oversized patient compared with the small patient (p < 0.001). As the size of the simulated patient increased, the radiation dose to the skin yielded a greater percentage increase than did the radiation dose to the abdominal organs (Fig. 4A, 4B).

Discussion

With high-output CT scanners with automatic tube current modulation technique, radiologists can obtain images with constant image quality despite the size of the patient. Because it is clinically desirable to maintain image quality over a range of patient sizes, radiologists are tempted to use a constant operator-selected image quality setting (e.g., noise index, reference tube current–time product) for patients of all sizes. From a patient dose perspective, however, this approach is not necessarily recommended. With a pre-selected image quality setting, a thin patient can receive a very low radiation dose and an oversized patient, a very high dose. To the best of our knowledge, the relation between radiation dose and patient size with the use of automatic tube current modulation in abdominal CT has not been investigated in an anthropometric phantom study.
The CT scanner used in our study delivered a maximum tube current of 715 mA at a peak tube potential of 140 kVp. Because of the high tube current capability of the scanner, the selected noise indexes, and the adjusted gantry rotation times, xyz-axis automatic tube current modulation maintained image quality for the three simulated patient sizes. However, the maintenance of constant image quality for larger patients came at the cost of significantly increased radiation. For the simulated oversized patient, the abdominal organ doses increased up to 528.1% and the skin doses up to 816.3%, compared with the doses in small patients. Thus when automatic tube current modulation is used without adjustment of the noise index to patient habitus, radiologists and technologists need to be aware of the resultant very broad range of radiation doses.
Fig. 4A Mean radiation dose for skin (+), spleen (▴), stomach (▪), and liver (♦) plotted according to size of simulated patient (lateral diameter). Graphs show doses for protocol A (A) and protocol B (B).
Fig. 4B Mean radiation dose for skin (+), spleen (▴), stomach (▪), and liver (♦) plotted according to size of simulated patient (lateral diameter). Graphs show doses for protocol A (A) and protocol B (B).
Operator-selected image quality settings play a key role in the dose efficiency of automatic tube current modulation. Previous studies [1, 11] have shown that subjective image quality increases with patient size on abdominopelvic CT images with constant image noise. In other words, radiologists tend to accept greater subjective image quality for larger patients. The increased subjective image quality for larger patients may be explained by the increase in fat deposition around the abdominal organs that adds to the inherent tissue contrast between various organs. This greater acceptance of higher subjective image quality for large patients may allow adjustment of the operator-selected image quality settings in the use of automatic tube current modulation. The optimal noise indexes adjusted to patient habitus for abdominopelvic CT still have to be evaluated in a clinical study [1].
Selection of image quality for different-sized patients undergoing abdominopelvic CT with automatic tube current modulation is widely operator dependent and thus arbitrary. No clear reference values for patients of various sizes have been published, to our knowledge. In an attempt to reduce the arbitrary nature of parameter selection, manufacturers of CT scanners should be encouraged to intensify research on semiautomatic selection of image quality settings dependent on patient size. Udayasankar et al. [12] assessed a sized-compensated tube current modulation technique using the squareroot projection area of the patient's scout image to select the noise index for abdominopelvic CT. The square root of the projection area can be determined from scout or CT images and is the most appropriate patient size metric for patient-dependent protocol planning because it is a direct measure of patient attenuation and not merely a factor for attenuation, as are patient dimensions, weight, and body mass index. In the study by Udayasankar et al., the preliminary results of size compensation were promising. Two radiologists graded the image quality of all 100 abdominopelvic CT examinations as acceptable regardless of patient size. The applied noise indexes used for the 100 CT examinations ranged between 9.6 and 28.7 H.
Another noteworthy finding of our study was the greater percentage increase in skin dose compared with abdominal organ dose as patient size increased. A possible explanation for this finding may be the greater attenuation of the traversing photon fluence through the wider girth of our simulated patient, resulting in lower detection of radiation by the abdominal detectors in the center of the phantom compared with the skin detectors on the body surface. If one assumes that the female breast can receive doses similar to our reported skin doses, oversized women are at risk of receiving high radiation doses to the radiosensitive breast with automatic tube current modulation CT. Further studies of this technique are needed to assess exposure of the female breast to radiation in different-sized patients.
There were limitations to our study. First, the two fat rings covered only 10 cm of the upper abdomen instead of the entire abdomen and pelvis. Total coverage of the abdomen and pelvis was not technically feasible because of the curvature of the phantom at the level of the lower abdomen and pelvis. A complete abdominopelvic CT examination of the three phantom setups with the two protocols should increase the absolute organ and skin dose numbers but not the percentage change between the phantom setups. Second, we used a female phantom and did not include a male phantom. The fat deposits of obese men tend to be mainly in the visceral region; in obese women, fat is present primarily in the subcutaneous region [13, 14]. The female phantom, however, which had a smaller cross-sectional diameter than that of a male phantom and did not have extra visceral fat deposition, simulates the worst-case radiation exposure scenario for a large patient. Furthermore, because our simulation included an increase in subcutaneous fat only and not an increase in visceral fat, our abdominal organ radiation dose results have to be considered higher than those for an actual obese patient of corresponding size. Third, we evaluated automatic tube current modulation with the xyz-axis modulation technique and did not separately evaluate z-axis and xy-axis modulation techniques. Because clinical studies [2, 3] have shown extra dose savings with the combined technique, we decided to investigate only the technique with the greater potential for dose reduction.
With abdominal CT performed with xyz-axis automatic tube current modulation, our phantom study showed a wide range of radiation doses due to varying patient size. As the patient's body habitus increased, the greatest increase in radiation exposure was measured in the skin. To ensure dose efficiency with automatic tube current modulation in patients of all sizes, we recommend that the operator-selected image quality settings be adjusted to the patient's habitus.

Footnotes

Address correspondence to R. C. Nelson ([email protected]).
WEB
This is a Web exclusive article.

References

1.
Kalra MK, Maher MM, Kamath RS, et al. Sixteen-detector row CT of abdomen and pelvis: study for optimization of Z-axis modulation technique performed in 153 patients. Radiology 2004; 233:241 –249
2.
Mulkens TH, Bellinck P, Baeyaert M, et al. Use of an automatic exposure control mechanism for dose optimization in multi-detector row CT examinations: clinical evaluation. Radiology 2005; 237:213 –223
3.
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; 186:673 –679
4.
Mulkens TH, Daineffe S, De Wijngaert R, et al. Urinary stone disease: comparison of standard-dose and low-dose with 4D MDCT tube current modulation. AJR 2007; 188:553–562
5.
Kalra MK, Maher MM, Toth TL, Kamath RS, Halpern EF, Saini S. Comparison of Z-axis automatic tube current modulation technique with fixed tube current CT scanning of abdomen and pelvis. Radiology 2004; 232:347–353
6.
Kalra MK, Maher MM, Toth TL, et al. Strategies for CT radiation dose optimization. Radiology 2004; 230:619–628
7.
McCollough CH, Bruesewitz MR, Kofler JM Jr. CT dose reduction and dose management tools: overview of available options. RadioGraphics 2006; 26:503–512
8.
Ogden K, Huda W, Scalzetti EM, Roskopf ML. Patient size and x-ray transmission in body CT. Health Phys 2004; 86:397 –405
9.
Hendee WR. Interaction of x and gamma rays in the body. In: Hendee, WR, ed. Medical radiation physics, 2nd ed, Chicago, IL: Year Book, 1979:171 –172
10.
Hurwitz LM, Yoshizumi TT, Goodman PC, et al. Effective dose determination using an anthropomorphic phantom and metal oxide semiconductor field effect transistor technology for clinical adult body multidetector array computed tomography protocols. J Comput Assist Tomogr 2007; 31:544 –549
11.
Wilting JE, Zwartkruis A, van Leeuwen MS, Timmer J, Kamphuis AG, Feldberg M. A rational approach to dose reduction in CT: individualized scan protocols. Eur Radiol 2001; 11:2627 –2632
12.
Udayasankar UK, Kalra M, Li J, et al. Multidetector scanning of the abdomen and pelvis: a study for evaluation of size compensated automatic tube current modulation technique in 100 subjects. RSNA 2006. Oak Brook, IL: Radiological Society of North America, 2006. Available at: http://rsna2006.rsna.org/rsna2006/V2006/conference/event_display.cfm?em_id=4441276. Accessed November 14, 2007
13.
Kanehisa H, Miyatani M, Azuma K, Kuno S, Fukunaga T. Influences of age and sex on abdominal muscle and subcutaneous fat thickness. Eur J Appl Physiol 2004; 91:534 –537
14.
Snehalatha C, Ramachandran A, Satyavani K, Vallabi MY, Viswanathan V. Computed axial tomographic scan measurement of abdominal fat distribution and its correlation with anthropometry and insulin secretion in healthy Asian Indians. Metabolism 1997; 46:1220 –1224

Information & Authors

Information

Published In

American Journal of Roentgenology
Pages: W100 - W105
PubMed: 18212190

History

Submitted: July 18, 2007
Accepted: September 10, 2007

Keywords

  1. automatic tube current modulation
  2. CT
  3. image quality
  4. obesity
  5. radiation dose

Authors

Affiliations

Sebastian T. Schindera
Department of Diagnostic Radiology, University Hospital of Bern, Bern, Switzerland.
Department of Radiology, Duke University Medical Center, Box 3808, Erwin Rd., Durham, NC 27710.
Rendon C. Nelson
Department of Radiology, Duke University Medical Center, Box 3808, Erwin Rd., Durham, NC 27710.
Thomas L. Toth
GE Healthcare Inc., Waukesha, WI.
Giao T. Nguyen
Division of Radiation Safety, Duke University Medical Center, Durham, NC.
Greta I. Toncheva
Division of Radiation Safety, Duke University Medical Center, Durham, NC.
David M. DeLong
Department of Radiology, Duke University Medical Center, Box 3808, Erwin Rd., Durham, NC 27710.
Terry T. Yoshizumi
Department of Radiology, Duke University Medical Center, Box 3808, Erwin Rd., Durham, NC 27710.
Division of Radiation Safety, Duke University Medical Center, Durham, NC.

Metrics & Citations

Metrics

Citations

Export Citations

To download the citation to this article, select your reference manager software.

Articles citing this article

View Options

View options

PDF

View PDF

PDF Download

Download PDF

Media

Figures

Other

Tables

Share

Share

Copy the content Link

Share on social media