Gastrointestinal Imaging Original Research

Transcription

1 Gastrointestinal Imaging Original Research Mayer et al. Radiation Dose Savings in CT Gastrointestinal Imaging Original Research Caroline Mayer 1 Mathias Meyer 1 Christian Fink 1 Bernhard Schmidt 2 Martin Sedlmair 2 Stefan O. Schoenberg 1 Thomas Henzler 1 Mayer C, Meyer M, Fink C, et al. Keywords: automatic tube current modulation, automatic tube voltage selection, CT, image quality, radiation dose reduction DOI:.2214/AJR Received July 29, 13; accepted after revision January 3, 14. C. Mayer and M. Meyer contributed equally to this work. B. Schmidt and M. Sedlmair are employees of Siemens Healthcare. The other authors (who have no conflicts of interest to disclose) had control of the data and information submitted for publication. 1 Institute of Clinical Radiology and Nuclear Medicine, University Medical Center Mannheim, Medical Faculty Mannheim - Heidelberg University, Theodor-Kutzer-Ufer 1-3, D Mannheim, Germany. Address correspondence to T. Henzler (thomas.henzler@umm.de). 2 CT Division, Siemens Healthcare Sector, Forchheim, Germany. AJR 14; 3: X/14/ American Roentgen Ray Society Potential for Radiation Dose Savings in Abdominal and Chest CT Using Automatic Tube Voltage Selection in Combination With Automatic Tube Current Modulation OBJECTIVE. The purpose of this study was to evaluate the simultaneous use of automatic tube current modulation (ATCM) and automatic tube voltage selection (ATVS) for abdominal and chest CT examinations regarding radiation dose reduction and image quality. MATERIALS AND METHODS. We enrolled 617 patients who all underwent contrastenhanced chest or abdominal CT and divided them into two groups. In group A, 317 patients who underwent CT with only ATCM and a fixed body mass index adjusted tube voltage (1 kv or 0 kv) were enrolled. In group B, both ATCM and ATVS were used. Image attenuation and noise were measured in different anatomic regions. RESULTS. The mean contrast-to-noise ratio and the signal-to-noise ratio of abdomen and chest CT was higher in group B compared with group A (p < ). In total, the effective radiation doses for abdomen and chest CT examinations were significantly reduced in group B by 18% compared with group A (p < ). When only examining those who benefited from the ATVS tool, a dose reduction of 35% for chest CT and 42% for abdomen CT could be achieved (p < for each). CONCLUSION. The simultaneous use of ATVS and ATCM enables significant radiation dose reduction in abdominal and thoracic contrast-enhanced CT examinations compared with the use of ATCM alone. S ince the introduction of MDCT, the total number and clinical indications for MDCT examinations has grown steadily. CT is by far the largest contributor to medical radiation exposure among the U.S. and European population [1]. This has raised significant concerns about potential radiation hazards from diagnostic CT. Although the risks for a single person undergoing a single examination may be neglected, the increasing radiation exposure in the population is going to play a role in future public health issues [2]. Consequently, radiologists, physicists, and manufacturers are trying to decrease radiation exposure to fulfill the principle of keeping radiation dose as low as reasonably achievable [3]. Within the past years, various techniques for radiation dose reduction have been developed, such as automatic tube current modulation (ATCM), reduced tube voltage based on patient size, cone beam bowtie filters, iterative reconstruction algorithms, and decreased scanning length [4]. ATCM, which has been introduced into clinical routine imaging, enables automatic adjustment of the tube current in various planes (x, y, or z) on the basis of the size and attenuation of the body area scanned, allowing a constant image quality and improved radiation dose efficiency [5, 6]. However, most of these approaches lead to a simultaneous increase in image noise and a decrease in acceptable images of diagnostic quality, limiting their potential mainly because of the higher absorption of low-energy photons by the patient [7]. Recently, attention has been shifted toward the potential of low-tube-voltage imaging and newly introduced automatic tube voltage selection (ATVS) in addition to ATCM [8 ]. ATVS automatically maintains a constant contrast-to-noise ratio (CNR) depending on the examination type (e.g., angiographic, parenchymal with contrast administration, or soft tissue without contrast administration). This is accomplished by transforming the scanning parameters, e.g., tube voltage (kv) and tube current (mas), from a predefined reference patient to the parameters of the selected examination type also respecting the deviation from the reference patient. 292 AJR:3, August 14

2 Radiation Dose Savings in CT The deviation is calculated using the attenuation information that is based on the initial topogram scan. Transforming the scanning parameters in general results in a reduction of overall patient dose because in most cases the tube voltage is reduced while the tube current is increased, leading to an overall reduction in patient dose. If the examination requires more tube current to maintain the CNR, the patient dose might increase in some cases. The aim of this study was to evaluate the simultaneous use of ATCM and ATVS for abdomen and thorax contrast-enhanced CT examinations regarding radiation dose reduction and image quality. Materials and Methods Patients This retrospective single-center study was approved by our institutional review board and was in compliance with the Declaration of Helsinki and HIPAA. Because of the retrospective nature of our study, informed consent for the retrospective data analysis was waived. All CT examinations were clinically indicated for the assessment of abdominal and thoracic abnormalities. In total, 617 consecutive adult patients were enrolled in this study and divided into two groups. The first group (group A) included 317 patients (mean age [± SD] 62 ± 16 years; 146 women and 171 men) who underwent a standard protocol of the abdomen (n = 158) or the chest (n = 159) using ATCM (CARE Dose4D, Siemens Healthcare) and a fixed tube voltage of 1 kv. The second group (group B) consisted of 0 patients (mean age, 64 ± 12 years; 123 women and 177 men) who underwent a modified CT protocol of the abdomen (n = 135) or the chest (n = 165) using ATCM in combination with ATVS (CARE kv, Siemens Healthcare). For both groups, the topogram scan was obtained with a fixed tube voltage of 1 kv. For quality assurance of this new tool, the data acquisition was divided into two steps for group B. In an initial step, the ATVS tool was turned off, leaving only the ATCM tool as the dose modulator. The system then generated a topogram, calculating the estimated scanning and dose parameters, volume CT dose index ( ), dose-length product (DLP), effective tube current (mas), and kilovoltage. These parameters were recorded for all patients because they reflect the scanning parameters of each patient as if the patient had been scanned with the Care kv tool. This step did not include any additional patient scanning. In a second step, ATVS was then turned on and the patient was scanned using ATCM and ATVS. CT Technique All CT examinations were performed on a dual-source 64-MDCT system (Somatom Definition, Siemens Healthcare). CT examinations were performed with two protocols. The first protocol (group A) was performed with a standard 1 kv and ATCM. The second protocol (group B) was performed using ATCM and ATVS simultaneously. Detector collimation was set to mm and a pitch of 1.2 was used for all abdominal and thoracic examinations. For the abdominal portal venous examinations, contrast enhancement was achieved by injecting 5 ml of nonionic iodinated contrast material (0 mg I/mL iomeprol, Iomeron, Bracco) via an antecubital vein at a flow rate of 2.5 ml/s using a dual-syringe power injector (Stellant D CT Injection System, Medrad). The examination automatically started after a fixed delay of 70 seconds after the initiation of contrast material injection. Contrast enhancement for the arterial thoracic examinations was performed by injecting 80 ml of nonionic iodinated contrast material (0 mg I/mL iomeprol) via an antecubital vein at a flow rate of 3.5 ml/s using a dual-syringe power injector (Stellant D CT Injection System). Scanning initiation was determined individually using bolus tracking with a triggered threshold of 0 HU within an ROI that was placed in the descending aorta. CT raw data were reconstructed using filtered back projection with a slice thickness of 1.5 mm using a soft-tissue convolution kernel (Bf). Principle of the Automatic Tube Voltage Selection Tool The ATVS tool that was evaluated in this study was designed to work in conjunction with ATCM, which modulates tube current on the basis of the topogram after the optimal tube voltage is selected by ATVS. According to the attenuation along the z-axis obtained from the CT topogram, the software algorithm automatically calculates the tube current for all possible kilovoltage (70 1 kv) values at each scanning location that meets the user-prescribed image quality requirements. A tube current modulation curve is calculated over the entire scanning range, and the radiation dose is estimated on the basis of the kv / mas curves. Some tube voltages might result in lower doses but are rejected by ATVS because of tube current limits (i.e., the required tube current will exceed the scanner s maximum tube power capacity). If the required tube current at any location exceeds the system limits, the calculation is repeated automatically for the next tube potential that yields optimal image quality and the lowest individual dose. The kilovoltage selected by ATVS is a single discrete selection, so that the same kilovoltage is used throughout the CT examination. Before the examination is performed with ATVS, the optimized kilovoltage, optimized effective tube current, and that the system plans to deliver are displayed on the scanner interface [8]. Quantitative Image Analysis All CT measurements were performed on an offline communication workstation (Osirix, version 3.7.1, Aycan) by a third-year resident with 2 years of experience in whole-body CT who was blinded to the CT protocol. CT values (HU) and image noise (SD of measured CT values) were measured on the reconstructed 1.5-mm axial CT images using equal ROIs in various anatomic regions. On the abdominal images, ROIs were placed within the liver, spleen, portal vein, abdominal aorta, psoas muscle, visceral fat, and air (Fig.1). On the thoracic images ROIs were placed within the descending aorta; subscapularis muscle; and air within the trachea, lung parenchyma, and subcutaneous fat. The ROIs were drawn by hand in the form of ovals of various sizes depending on the anatomic region. However, the size in one anatomic region was kept equal throughout the patient evaluations. Each ROI was placed in the identical or nearly identical segment. SDs of the attenuation in these ROIs were recorded. Each anatomic area was measured three times, and the mean value of these three evaluations was used for further analysis. Subsequently, signal-to-noise ratio (SNR) and CNR were calculated for each image dataset according to the methods used by Szucs-Farkas et al. [11]. In detail, the following formulas were applied: SNR = mean HU measured in ROI / noise (SD from mean attenuation measured in ROI). For example, for the chest: mean attenuation in ROI of the descending aorta / SD from mean attenuation measured in ROI of the descending aorta, and CNR = mean HU measured in ROI mean HU measured in ROI of the surrounding tissue / SD from mean attenuation measured in ROI. For example, for the abdomen: mean HU measured in ROI of the liver mean HU measured in ROI of the subscapularis muscle) / SD from mean attenuation measured in ROI of the liver. Qualitative Image Analysis Subjective image quality was initially assessed separately by two radiologists with 2 and 6 years of experience and later in a consensus reading. Both radiologists rated the assessability of thoracic and abdominal structures, image noise, and diagnostic acceptability as described for wholebody staging CT examinations in the European Guidelines on Quality Criteria for CT [12]. Images were evaluated for overall diagnostic image quality on a 5-point Likert scale (1, poor; 2, fair; AJR:3, August

3 Mayer et al. 3, moderate; 4, good; and 5, excellent). Additionally, image noise was subjectively evaluated using the following 5-point Likert scale (1, unacceptable image noise; 2, above average noise; 3, average image noise; 4, less than average noise; and 5, minimal image noise). A C Fig. 1 Comparison of patient groups. A D, Axial contrast-enhanced CT images of 67-year-old man from group B (A and B) and 55-year-old man from group A (C and D). During atrial phase (A) and portal venous phase (B), images were obtained for combined approach in group B at tube voltage of 0 kv. Mean image noise exemplarily measured in portal vein (white circle) and latissimus dorsi muscle (black circle) was 24 and 23 HU (A) and 25 and 22 HU (B). In group A with fixed tube voltage of 1 kv exemplarily measured mean image noise was 24 and 18 HU during atrial phase (C) and 24 and 21 HU during portal venous phase (D). Radiation Dose Assessment The, DLP, effective diameter, and size-specific estimate were obtained from the CT dose report for each patient for the abdomen and the chest acquisitions [13]. The effective dose was estimated by multiplying the DLP by body-specific conversion coefficients according to the generic method presented in the European Guidelines for MDCT (dose conversion coefficients in msv Gy 1 cm 1 : thorax, and abdomen-pelvis, 0.017) [14]. To detect any relationship between body constitution and radiation dose, we evaluated body habitus by taking the maximum anteroposterior and the transverse diameters of the abdomen and the chest. Statistical Analysis All statistical analyses were performed using SAS 9.0 software (SAS Institute) at a significance level of Continuous variables are expressed as mean ± SD and categoric variables as percentages. To compare the CT density and image noise, SNR, CNR, and radiation parameters between both groups one-way ANOVA was performed. To compare the relationship between body constitution and radiation dose, Pearson correlation was performed. Results Body Constitution No statistically significant differences within the measured body constitution parameters were found between both groups, in particular, for the chest anteroposterior diameter (group A, 23.2 ± 2.9 cm vs group B, 23.7 ± 2.6 cm; p > 0.05) and transverse diameter (group A, 33.0 ± 4.3 cm vs group B, 34.7 ± 4.3 cm; p > 0.05) and for the abdomen anteroposterior diameter (group A, B D 294 AJR:3, August 14

4 Radiation Dose Savings in CT TABLE 1: Radiation Dose Parameters of Group A and Group B Group A (n = 317) Group B (n = 0) p (ATCM vs Simultaneous Use of ATCM and ATVS) Radiation Dose Parameters 26.2 ± 4.1 cm vs group B, 25.3 ± 3.5 cm) and transverse diameter (group A, 33.2 ± 3.5 cm vs group B, 32.6 ± 3.1 cm; p > 0.05 for both). Radiation Dose The overall comparison of both groups showed a significant dose reduction of 18% in group B compared with group A (p < ). No significant differences in tube current modulation were observed in group A and group B. Radiation dose for groups A and B as well as the estimated dose parameters for the single use of ATCM in group B are displayed in Table 1. Chest CT Among the 165 chest examinations in group B, tube voltage was reduced in 96 patients from 1 to 0 kv and in three patients to 80 kv (99 patients; 61%) (Table 2). In total an overall dose reduction of 16.8% was found for group B compared with group A (p = ) (Table 1). In a second step, we performed a subgroup analysis by separating patients of group B who benefited from the ATVS tool because the tube voltage of their examination was reduced because of this program. This was the case for 99 patients undergoing chest CT and for 48 patients undergoing abdominal CT. The radiation dose parameters were assessed and compared with the parameters of group A. Comparing this subgroup to group A, a dose reduction of 34% could be achieved (mean DLP, ± ATCM Only ATCM Only ATCM and ATVS Within Group B Group A vs Group B Chest DLP (mgy cm ) ± ± ± (mgy) 7.9 ± ± ± Effective current time product per tube (mas) ± ± ± Effective patient diameter (cm) 27.6 ± ± Size-specific estimate (mgy).4 ± ± ± Effective dose (msv) 6.0 ± ± ± Abdomen DLP (mgy cm ) ± ± ± (mgy) 12.3 ± ± ± Effective current time product per tube (mas) ± ± ± Effective patient diameter (cm) 29.5 ± ± Size-specific estimate (mgy) 15.1 ± ± ± Effective dose (msv).7 ± ± ± Note Except for p values, data are mean ± SD. ATCM = automatic tube current modulation, ATVS = automatic tube voltage selection, DLP = dose-length product, = volume CT dose index. Dash indicates not applicable. mgy for group A and mean DLP, 6.5 ± 49 mgy for subgroup B; p < ). Furthermore, no differences in tube current modulation were observed between both groups (group A, ± 66.6 mas vs subgroup B, ± 26.7 mas; p = ). To reduce bias due to intergroup differences, we also compared the estimated data of an ATCM single use in group B with the true Care kv data for group B. By comparing the dose parameters of these two groups, a dose reduction of 19% could be achieved (mean DLP, ± 0.3 mgy vs ± 97.8 mgy; p < ) (Table 1). The quantitative image analysis revealed no significant differences in image noise for chest CT between both groups, with the exception of the lung (p = ) (Fig. 2). Mean density was significantly higher in fat tissue and the ascending aorta for group B compared with group A (all p < ) (Table 3). In addition, the CNR and the SNR of the ascending aorta were significantly higher in group B (both p < ). A detailed list of the quantitative values is displayed in Table 3. Independent of the approach chosen, all images were rated as diagnostically sufficient, and the level of interobserver agreement was excellent (for overall image quality, κ = 0.9, and for image noise, κ = 0.84). The median range image quality scores (5 [4 5] vs 5 [4 5]; p = 0.816) as well as the TABLE 2: Recommended Kilovoltage Settings for Group B Tube Voltage (kv) ATCM Only ATCM and ATVS Chest Abdomen Note Data are number of examinations. ATCM = automatic tube current modulation, ATVS = automatic tube voltage selection. image noise scores (5 [4 5] vs 5 [4 5]; p = 0.721) assigned by the two radiologists did not differ significantly between the two groups for chest CT. Abdominal CT Among the 135 abdominal CT examinations in group B, tube potential was reduced to 0 kv in 48, increased to 1 kv in four, and remained unchanged in 83 (Table 2). This resulted in a dose reduction of 18.4% for group B compared with the dose parameters for group AJR:3, August

5 Mayer et al. Mean Image Noise per Region (HU) Group A Abdominal aorta Air Fat Liver Group B Psoas muscle Portal vein Spleen Group A TABLE 3: Quantitative Analysis Results in Group A and Group B A Parameter Group A (n = 317) Group B (n = 0) p Chest and abdomen density values (HU) Descending aorta ± ± Subscapularis muscle 45.8 ± ± Trachea 925 ± ± Lung 842 ± ± Chest fat 96.6 ± ± Liver 84.4 ± ± Spleen 94.7 ± ± Portal vein ± ± Abdominal aorta ± ± Psoas muscle 48 ± ± Abdominal fat 98.8 ± ± Air ± ± Chest and abdomen CNR values Ascending aorta 5.8 ± ± Liver 2.0 ± ± Portal vein 3.6 ± ± Abdominal aorta 3.2 ± ± Chest and abdomen SNR values Descending aorta 8.1 ± ± Trachea 28.6 ± ± Lung.7 ± ± Chest fat 5.1 ± ± Liver 4.6 ± ± Portal vein 5.6 ± ± Abdominal aorta 5.1 ± ± Abdominal fat 6.4 ± ± Note CNR = contrast-to-noise ratio, SNR = signal-to-noise ratio. Mean Image Noise per Region (HU) Ascending aorta Fat Lung Group B Subscapularis muscle Trachea B Fig. 2 Mean noise levels for soft-tissue regions. A and B, Graphs show plots of mean noise levels for soft-tissue regions determined for two different groups investigated in this study (group A, automatic tube current modulation only and group B, automatic tube current modulation and tube voltage selection) for abdomen (A) and chest (B). Note that some data points overlap. Horizontal lines represent mean image noise per reconstruction technique. A (p = 0.016) (Table 1). Comparing the parameters of group A with those who benefited from Care kv in group B (48 patients), as the tube voltage was reduced, an even higher dose reduction of 42% could be achieved (mean DLP group A, ± mgy vs group B, 364 ± 79.7 mgy; p < ). When again looking at the estimated dose parameters of the ATCM single examination in group B and comparing it with the true Care kv data for group B, we found a dose reduction of 17% (mean DLP, ± mgy vs ± mgy; p < ) (Table 1). A significant moderate linear correlation was observed between body constitution and for chest CT (anteroposterior diameter r = 0.46 and transverse diameter r = 0.55, p < ) and for the DLP (anteroposterior diameter r = 0.45 and transverse diameter r = 0.51; p < ). For the abdominal CT examinations, the correlation between body constitution parameters and / DLP was lower when compared with the chest examinations but still statistically significant ( : anteroposterior diameter r = 0.32 and transverse diameter r = 0.36, p < ; DLP: anteroposterior diameter r = 0.28 and transverse diameter r = 0.29, p < ). The results are shown in Figure 3. The quantitative image analysis revealed that image noise was significantly higher in group B within the spleen, liver, and fat tissue compared with group A (p < ) (Fig. 1). The mean density values for group B were only significantly higher in the abdominal muscle tissue compared with group A. In addition, the liver 296 AJR:3, August 14

6 Radiation Dose Savings in CT (mgy) (mgy) 50 r = 0.46 p < Thorax Anteroposterior (cm) r = 0.32 p < Thorax Anteroposterior (cm) CNR was significantly higher in group B (p = ), whereas no significant difference in CNR was observed for all other organs. Furthermore, the SNR was significantly higher in the abdominal fat tissue of group B (p < ), whereas significantly lower SNR values were found in the liver parenchyma compared with group A (p = ). The values for objective image quality for both groups are summarized in Table 3. Independent of the approach chosen, all images were rated as diagnostically sufficient, and the level of interobserver agreement was excellent (for overall image quality, κ = 0.89 and for image noise, κ = 0.81). The median range image quality scores (5 [4 5] vs 5 [4 5]; p = 0.645) as well as the image noise scores (5 [4 5] vs 5 [4 5]; p = 0.567) assigned by the two radiologists did not differ significantly between the two groups for abdominal CT. A (mgy) r = 0.55 p < Thorax Transverse (cm) Thorax Transverse (cm) C D Fig. 3 Correlation between patient measurement and volume CT dose index ( ). A D, Graphs show linear statistical correlation between patient anteroposterior and transverse body diameter and volume in abdominal and chest CT: A, anteroposterior chest CT; B, transverse chest CT; C, anteroposterior abdominal CT; and D, transverse abdominal CT. (mgy) 50 r = 0.36 p < Discussion There are many CT technical parameters that can be adjusted to reduce radiation dose on CT, including tube potential; tube current; iterative reconstruction algorithms; and, with scanners from some manufactures, the pitch [15 17]. However, adjusting one parameter independently of all others will increase image noise and may compromise image quality, with the exception of iterative reconstruction algorithms, which make it possible to diminish this problem [5, 18]. Among the various dose-reduction techniques, ATCM, which automatically adjusts the tube current on the basis of patient size and attenuation, has become one of the most comprehensive approaches and is therefore used in clinical routine [6]. Several studies have evaluated this technique in different body regions, including the head, neck, chest, and abdomen. All of these studies concluded that a significant dose reduction between 15% and 50% could be achieved while maintaining sufficient image quality by using ATCM techniques [5, 6, 19]. Given that tube voltage has an exponential relationship with radiation dose [], lowering the tube voltage may also be an effective approach to reduce radiation exposure because of the introduction of ATVS, which recommends the optimal B tube current and tube voltage settings for each patient on the basis of body constitution and body attenuation obtained from the individual CT topogram. Hence, the optimal combination of tube voltage and tube current modulation is calculated, resulting in the lowest achievable radiation dose for each patient. Several studies have evaluated ATVS in phantom- and patientbased CT angiography (CTA) and abdominal CT studies [8,, 21]. Lee et al. [8] assessed dose reduction and image quality of contrast-enhanced liver CT using ATVS. The authors observed an overall dose reduction of 31% depending on the contrast gain settings using ATVS, which is in accordance to the level of dose reduction found in our study for abdominal CT. The overall image quality was acceptable despite higher image noise in ATVS studies. They also reported a high correlation between the patient s BMI and dose reduction, noting that underweight patients had more benefits of dose reduction (33%) than normal weight patients [8]. Goetti et al. [] examined the impact of automatic tube potential selection on image quality and radiation dose in 35 patients undergoing whole-body CTA after endovascular aneurysm repair of the abdominal aorta. They showed that ATVS led to a decrease of tube voltage to 0 and 80 kv in 51% of their patients and an overall radiation dose reduction of 16% compared with imaging with a fixed tube voltage of 1 kv. Image noise was higher in the ATVS group as well but still allowed sufficient diagnostic quality []. Hough et al. [21] reported similar results. This group compared ATVS in 36 patients undergoing contrast-enhanced abdominopelvic CT with a size-matched control group. In this retrospective study, a dose reduction between 18% and 25% depending on the tube voltage setting (1 or 0 kv) was achieved [21]. Our study showed that by using ATVS an overall decrease in radiation dose of 18% is feasible. In particular for chest and abdomen CT, dose reductions of 16.8% and 18.4%, respectively, were observed compared with the ATCM alone. These results are in accordance with the previously mentioned results showing the feasibility of this tool for chest and abdominal CT. In comparison with previous studies, we further analyzed those patients who benefited from ATVS and preformed a subgroup analysis. In 35% of the abdomen examinations and in 60% of the chest examinations, tube voltage was reduced to AJR:3, August

7 Mayer et al. the next lower level. For these particular patients, an overall dose reduction of 42% in abdomen CT and 34% in chest CT could be achieved compared with a single ATCM approach. Surprisingly, tube current modulation did not increase simultaneously with a decrease in tube voltage on the ATVS study as we expected. The tube current modulation was similar in both groups. Usually lowering the tube voltage leads to a compensatory increase in tube current because both parameters depend on each other. In contrast to the previously mentioned studies, we did not use BMI but measured body anteroposterior and transverse diameters instead. As shown in several other studies, BMI is not a reliable parameter to assess patient habitus. Over- or underdosing can occur when BMI is used to select tube potential [22]. In addition, reduced radiation dose and substantially improved image quality could be achieved when accounting for differences in body habitus and size according to anteroposterior and transverse diameters instead of BMI [23]. A significant weak to moderate linear correlation was observed between body diameter and radiation dose parameters. As shown in previous studies, patients with a small body diameter received lower radiation doses than those of with large body diameters [5, 8]. Regarding the quantitative analysis of our study, image noise and CNR increased significantly in some anatomic regions with ATVS, but the range of image noise was rated diagnostically acceptable. Whether the combined use of Care kv and newly reintroduced iterative reconstructions may diminish the increase in image noise has to be assessed in further studies. On the other hand, low-kilovoltage examinations provide higher vessel lumen attenuation of iodinated contrast media, which may be beneficial in detecting smaller differences in the vascularization of lesions because the lesion-to-background contrast increases [24]. Different studies have shown the benefit of this relation in the detection of small hypervascular lesions of the liver or pancreas [25, 26]. Because subjective image quality is particularly important to maintain diagnostic image quality when reducing the tube voltage setting, our results displayed no differences in subjective image quality or image noise regardless of the approach chosen. These findings are in accordance with previous studies that reported high diagnostic quality although tube voltage and tube modulation were reduced [, 21] Several limitations have to be acknowledged. First, our data may be influenced by an intergroup bias because patients and conditions were different between the two groups. To minimize this bias, we further compared the estimated data of using ATCM alone in group B with the true ATVS data for group B so that a comparison of both programs in the same patients would be possible. These results revealed findings similar to the comparison between group A and group B. Finally, the ATVS technology is vendor specific, which may limit its use in clinical practice. In conclusion, the use of ATVS for chest and abdominal CT results in a significant dose reduction while maintaining adequate image quality and diagnostic confidence without user interaction. The ATVS tool reduced effective tube voltage in many patients (49%), resulting in a dose reduction of 18%, showing the potential of this new dose modulation tool. Acknowledgments We thank Thomas Flohr and Paul Apfaltrer for their support. References 1. Mettler FA Jr, Bhargavan M, Faulkner K, et al. Radiologic and nuclear medicine studies in the United States and worldwide: frequency, radiation dose, and comparison with other radiation sources Radiology 09; 253: Brenner DJ, Hall EJ. Computed tomography: an increasing source of radiation exposure. N Engl J Med 07; 357: Mountford PJ, Temperton DH. Recommendations of the International Commission on Radiological Protection (ICRP) Eur J Nucl Med 1992; 19: May MS, Wust W, Brand M, et al. Dose reduction in abdominal computed tomography: intraindividual comparison of image quality of full-dose standard and half-dose iterative reconstructions with dual-source computed tomography. Invest Radiol 11; 46: Lee EJ, Lee SK, Agid R, Howard P, Bae JM, ter- Brugge K. Comparison of image quality and radiation dose between fixed tube current and combined automatic tube current modulation in craniocervical CT angiography. AJNR 09; : Lee S, Yoon SW, Yoo SM, et al. Comparison of image quality and radiation dose between combined automatic tube current modulation and fixed tube current technique in CT of abdomen and pelvis. Acta Radiol 11; 52: Yu L, Li H, Fletcher JG, McCollough CH. Automatic selection of tube potential for radiation dose reduction in CT: a general strategy. Med Phys ; 37: Lee KH, Lee JM, Moon SK, et al. Attenuationbased automatic tube voltage selection and tube current modulation for dose reduction at contrastenhanced liver CT. Radiology 12; 265: Winklehner A, Goetti R, Baumueller S, et al. Automated attenuation-based tube potential selection for thoracoabdominal computed tomography angiography: improved dose effectiveness. Invest Radiol 11; 46: Goetti R, Winklehner A, Gordic S, et al. Automated attenuation-based kilovoltage selection: preliminary observations in patients after endovascular aneurysm repair of the abdominal aorta. AJR 12; 199:[web]W380 W Szucs-Farkas Z, Strautz T, Patak MA, Kurmann L, Vock P, Schidera ST. Is body weight the most appropriate criterion to select patients eligible for low-dose pulmonary CT angiography? Analysis of objective and subjective image quality at 80 kvp in 0 patients. Eur Radiol 09; 19: Danish Society of Radiology website. European guidelines on quality criteria for computed tomography. Published Accessed April 15, American Association of Physicists in Medicine website. Size-specific dose estimates (SSDE) in pediatric and adult body CT examinations. Report 4. Published 11. Accessed April 15, [No authors listed]. The 07 Recommendations of the International Commission on Radiological Protection. ICRP publication 3. Ann ICRP 07; 37: McCollough CH, Bruesewitz MR, Kofler JM Jr. CT dose reduction and dose management tools: overview of available options. RadioGraphics 06; 26: Cody DD, Moxley DM, Krugh KT, O Daniel JC, Wagner LK, Eftekhari F. Strategies for formulating appropriate MDCT techniques when imaging the chest, abdomen, and pelvis in pediatric patients. AJR 04; 182: Huda W, Lieberman KA, Chang J, Roskopf ML. Patient size and x-ray technique factors in head computed tomography examinations. I. Radiation doses. Med Phys 04; 31: Kalra MK, Maher MM, Toth TL, et al. Techniques and applications of automatic tube current modulation for CT. Radiology 04; 233: Greess H, Normayr A, Wolf H, et al. Dose reduc- 298 AJR:3, August 14

8 Radiation Dose Savings in CT tion in CT examination of children by an attenuation-based on-line modulation of tube current (CARE Dose). Eur Radiol 02; 12: Gunn ML, Kohr JR. State of the art: technologies for computed tomography dose reduction. Emerg Radiol ; 17: Hough DM, Fletcher JG, Grant KL, et al. Lowering kilovoltage to reduce radiation dose in contrast-enhanced abdominal CT: initial assessment of a prototype automated kilovoltage selection tool. AJR 12; 199: Ghoshhajra BB, Engel LC, Major GP, et al. Direct chest area measurement: a potential anthropometric replacement for BMI to inform cardiac CT dose parameters? J Cardiovasc Comput Tomogr 11; 5: van der Wall EE, Schuijf JD, Bax JJ. Use of the anterior-posterior chest diameter in CT: reduction in radiation dose? Int J Cardiovasc Imaging ; 26: Eller A, May MS, Scharf M, et al. Attenuationbased automatic kilovolt selection in abdominal computed tomography: effects on radiation exposure and image quality. Invest Radiol 12; FOR YOUR INFORMATION Mark your calendar for the following ARRS annual meetings: April 19 24, 15 Toronto Convention Centre, Toronto, ON, Canada April 17 22, 16 Los Angeles Convention Center, Los Angeles, CA April May 5, 17 Hyatt Regency New Orleans, New Orleans, LA April 22 27, 18 Marriott Wardman Park Hotel, Washington DC 47: Marin D, Nelson RC, Samei E, et al. Hypervascular liver tumors: low tube voltage, high tube current multidetector CT during late hepatic arterial phase for detection initial clinical experience. Radiology 09; 251: Marin D, Nelson RC, Barnhart H, et al. Detection of pancreatic tumors, image quality, and radiation dose during the pancreatic parenchymal phase: effect of a low-tube-voltage, high-tube-current CT technique preliminary results. Radiology ; 256: AJR:3, August