CT Pulmonary Angiography Using Automatic Tube Current Modulation Combination with Different Noise Index with Iterative Reconstruction Algorithm in Different Body Mass Index: Image Quality and Radiation Dose

ARTICLE IN PRESS

Original Investigation

CT Pulmonary Angiography Using Automatic Tube Current Modulation Combination with Different Noise Index with Iterative Reconstruction Algorithm in Different Body Mass Index: Image Quality and Radiation Dose Yong-Xia Zhao, PhD, Zi-Wei Zuo, MM, Hong-Na Suo, MM, Jia-ning Wang, MD, Jin Chang, PhD, MD Rationale and Objectives: This study aimed to determine the appropriate body mass index (BMI)-dependent noise index (NI) setting in computed tomography pulmonary angiography (CTPA) with automatic tube current modulation with adaptive statistical iterative reconstruction (ASiR). Materials and Methods: A total of 480 patients who had a CTPA were divided into group A (18.5 kg/m2 ≤ BMI < 25 kg/m2), group B (25 kg/m2 ≤ BMI < 30 kg/m2), and group C (BMI ≥ 30 kg/m2), according to their BMI values; each group had 160 patients. The three groups were further randomly divided into four subgroups: A1, A2, A3, A4; B1, B2, B3, B4; and C1, C2, C3, C4, with corresponding NI values of 26, 36, 40, and 46, respectively. All images were restructured with the ASiR algorithm, and the images with the lowest NI (26 Hounsfield units) in each group were used as reference standard. The signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) for the pulmonary artery of each group were calculated. Subjective image quality was evaluated using a five-score method by two independent radiologists. The CT dose index of volume and dose-length product were recorded and were converted to effective dose (ED). SNR and CNR in the group A, B, and C subgroups were compared to repeated measures analysis of variance, and the subjective score, Volumetric CT dose index of volume, dose-length product, and ED were compared to one-way analysis of variance. Results: For groups A and B, the SNR, CNR, and subjective scores of the images in their subgroups showed no statistical differences (P > .05). The ED in subgroups A4 and B4 was significantly lower than that in subgroups A1 (by 33.24%) and B1 (by 34.47%) (P < .01). For group C, there was no significant difference in the SNR, CNR, and the subjective image scores between subgroups C3 and C1 (P > .05). The ED in subgroup C3 was significantly lower than the ED in subgroup C1 (by 47.75%) (P < .01) Conclusions: Patient BMI-dependent NI settings that are higher than the recommended value may be used in CTPA with automatic tube current modulation and ASiR to effectively reduce radiation dose while maintaining diagnostic image quality. Key Words: Automatic tube current modulation; noise index; radiation dose; image quality; CT pulmonary angiography. © 2016 The Association of University Radiologists. Published by Elsevier Inc. All rights reserved.

INTRODUCTION Acad Radiol 2016; ■:■■–■■ From the Department of Radiology, The Affiliated Hospital of Hebei University, 212 Yuhua East Road Baoding City Hebei Province, Baoding 071000, China (Y.-X.Z., Z.-W.Z., J.W.); Medicine School, Hebei University, Baoding 071000 (H.-N.S.); School of Life Science, Tianjin University, Tianjin 300072, China (J.C.). Received April 11, 2016; revised June 20, 2016; accepted July 1, 2016. Address correspondence to: Y.-X.Z. e-mail: [email protected] © 2016 The Association of University Radiologists. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.acra.2016.07.018

C

omputed tomography pulmonary angiography (CTPA) is currently a widely accessible and quick to perform technique for patients suspected of having pulmonary embolism (PE) in clinic (1,2), and it has a high sensitivity (94%–100%) and specificity (89%–100%) for the diagnosis of acute PE (3). However, the amount of radiation exposure to the patient population in CTPA and its risks have also increased concerns regarding potential radiation damage (4,5). Thus, reducing the radiation dose on CTPA 1

ZHAO ET AL

studies has become a priority for radiologists. For clinical populations in particular, it is important to reduce radiation dose by reducing the tube current (6). The automatic tube current modulation (ATCM) technique selects the optimal tube current in an automated manner by using the attenuation values on anteroposterior and lateral scanogram, and it is an important method to reduce radiation dose. In one implementation of ATCM technique, the noise index (NI) value can be used to control the output of the tube current, so the NI is used as an indirect representation of image quality. However, reducing the tube current can increase image noise and adversely affect image quality. Higher image noise can be reduced by different advanced reconstruction algorithms in maintaining low radiation doses; adaptive statistical iterative reconstruction (ASiR) is a widely used and effective way to reduce noise and maintain image quality (7–10). To our knowledge, no one has compared image quality and radiation dose in CTPA for the combination of different NI settings and different body mass index (BMI) values. The purpose of our study was to determine the appropriate BMI-dependent noise NI setting in CTPA with ATCM and ASiR algorithm by comparing radiation dose and image quality to the combination of different NI values and BMI values to maximize dose reduction while maintaining image quality. MATERIALS AND METHODS Patient Population

Between June 2014 and December 2015, 480 patients (248 men, 53.5 ± 20.5 years; 232 women, 54.7 ± 21.2 years) with suspected PE were included in the study, and indications for performing pulmonary computed tomographic angiography (CTA) were based on clinical assessment, abnormal findings at laboratory testing (blood gas analysis and plasmatic d-dimer level), abnormal results at echocardiography or electrocardiogram suggestive of acute right heart dysfunction, and findings suggestive of PE at conventional chest x-ray. Exclusion criteria were known allergy to iodinated contrast medium, glomerular filtration rate less than 60 mL/min, and hyperthyroidism. All patients had CTPA examination. The patients were divided into group A (18.5 kg/m2 ≤ BMI < 25 kg/m2), group B (25 kg/m 2 ≤ BMI < 30 kg/m 2 ), and group C (BMI ≥ 30 kg/m2), according to BMI; each group had 160 patients. These three groups were further randomly divided into four subgroups: A1, A2, A3, A4; B1, B2, B3, B4; and C1, C2, C3, C4. Each subgroup had 40 patients. Patient characteristics are shown in Table 1. CT Scan Protocols and Radiation Dose Estimation

All 480 patients underwent CTPA with the Gemstone detector to derive 128 slices per rotation on a high-definition Discovery CT750 HD (HDCT, GE Healthcare, Waukesha, WI). Patients were examined in the supine position during a single fullinspiration breath-hold with their arms raised behind their heads. 2

Academic Radiology, Vol ■, No ■■, ■■ 2016

TABLE 1. Patient Characteristics

Parameter

Group A (n = 160)

Patient characteristic 56.5 ± 17.8 Age (y)* Weight (kg)* 62.5 ± 8.1 Height (cm)* 169 ± 12.7 F/M 81/79 Axial length 313 (mm)*

Group B (n = 160)

Group C (n = 160)

P Value

57.7 ± 18.2 67.7 ± 7.8 168 ± 13.2 77/83 316

55.9 ± 17.6 79.5 ± 7.4 160 ± 14.5 74/86 310

.6567 .0971 .4157 .7344 .551

* Data are expressed as mean standard deviation (SD). P < .05 was considered statistically significant.

The CT protocols of all patients employed the ATCM technique. The technique adapts the tube current based on the patient’s attenuation at different projection angles in the x- and y-axes and along the z-axis at different table positions along the patient’s length. The technique is intended to achieve the desired image quality as specified by the user in the form of an NI with a user-selected minimum and maximum mA range. The scan parameters included helical, 0.5 second tube rotation time, pitch factor of 0.984:1, 50 cm field of view, table speed of 78.75 mm/s, tube voltage of 120 kVp, and tube current range of 30– 650 mA. The NI value in this study was gradually implemented; the recommended NI value of the CTPA by the manufacturer was 25. The image quality of 30 patients whose BMI was 18.5–33 kg/m2 and used this NI value was excellent, the NI value was set at 26 for patients in subgroups A1, B1, and C1. Because the image quality in the subgroups with NI = 26 was better than the diagnostic standard, NI = 36, 40, and 46 were respectively used with the patients in subgroups A2, B2, C2; A3, B3, C3; and A4, B4, C4 to investigate the dose reduction potential. Because of the fear of failure in the patient experiment, there was no use of NI value higher than 46. All the images were restructured with 50% ASiR (11); 50% ASiR setting implies 50% filtered back-projection (FBP) reconstruction blending with 50% ASiR in the reconstructed images. The scan coverage was from 1 cm lower than the clavicle to the diaphragm, and the axial length was from 280 mm to 330 mm. For all patients, the volumetric CT dose index (CTDIvol) and doselength product (DLP) were recorded from the scanner. The DLP was converted to effective dose (ED) in millisieverts (mSv); the ED was calculated according to the formula ED = DLP × 0.017, where 0.017 denotes the conversion factor (12). All the patients were injected with the contrast media, ioversol (Ioversolum 350, Hengrui Healthcare, China) through the median cubital vein. As HDCT iodine mapping is very sensitive for dense contrast media in the subclavian vein and superior vena cava, the use of a compact contrast media bolus with a good wash out of the venous system is essential to avoid artifacts. Therefore, we used a 45 mL bolus of pure contrast media, followed by 30 mL of a mixed phase with 40% of contrast media and 60% of NaCl, followed by a 50 mL chaser

Academic Radiology, Vol ■, No ■■, ■■ 2016

CTPA USING AUTOMATIC TUBE CURRENT MODULATION

Figure 1. Comparison of image quality. Images give an example of image impression achieved with automatic tube current modulation (ATCM) combination of noise index (NI) = 26 with a BMI of 23.32 kg/m2 (a), ATCM combination of NI = 36 with a BMI of 26.57 kg/m2 (b), ATCM combination of NI = 40 with a BMI of 32.21 kg/m2 (c), and ATCM combination of NI = 46 with a BMI of 31.35 kg/m2 (d). Section thickness is 0.625 mm for images a–d. Signal intensity was measured with a region of interest (ROI) tool as computed tomography (CT) value of main pulmonary and image noise as standard deviation of main pulmonary. Typical signal intensity values are noted with 373 Hounsfield units (HU) (a) and 363 HU (c) for 3D Smart with NI = 26 and 40. There was no significant difference between the two CT scan protocols (P > .05) with regard to signal intensity values and the image noise (17.41 and 17.02) (P > .05). There was a significant difference in signal intensity value between d (347 HU) and a (373 HU); d was lower than a by 7.3%.

bolus of pure NaCl in all groups. This protocol has shown satisfying results with vascular enhancement and low artifact burden on lung perfusion analysis of our daily routine use. In all patients, the timing of the scan was determined with smart prep technique, with a region of interest (ROI) placed on the main pulmonary artery, and scan was started 6 seconds after the ROI reached the threshold of 50 Hounsfield units. Objective Evaluation of Image Quality

The images of the patients were reconstructed by using a dedicated workstation. The thickness of the reconstructed images was 0.625 mm, at an interval of 0.6 mm. For the assessment of objective image quality, we measured the mean CT value and image noise in seven areas at the same level of axial images from each patient on the workstation. Circular ROIs with an area of 35 mm2 ± 2.0 mm2 were drawn in the main pulmonary trunk, right pulmonary artery, right pulmonary lobar arteries, left pulmonary artery, left pulmonary lobar arteries, the erector spine muscle, and the chest wall fat (Fig 1). Areas of focal changes in lung, ribs, esophagus, and airways, if any, were carefully avoided. For all measurements, the size, shape, and position of the ROIs were kept constant among the image sets by application of a copy-and-paste function in the workstation. The ROIs were positioned in homogeneous

portions of the structures. The attenuation was measured as the mean Hounsfield unit value in the ROI, and the noise was measured as standard deviations of the attenuation value of the ROI. The contrast-to-noise ratio (CNR) of all the images was calculated according to the formula (13,14): CNR = (CT main pulmonary trunk − CT erector spine muscle)/standard deviation chest wall fat. The signal-to-noise ratio (SNR) was calculated according to the formula (14,15) SNR = CT main pulmonary trunk/standard deviation main pulmonary trunk. The CNR and SNR measurements were obtained from the images for the objective evaluation of image quality. Subjective Evaluation of Image Quality

Image quality was further rated subjectively and independently by two blinded radiologists with more than 10 years of experience in chest imaging and CTA imaging, according to a five-point rating scale (16,17) (1 = unacceptable image quality, needed to be repeated; 2 = still diagnostic to lobar level but with significantly reduced confidence beyond; 3 = standard image quality, exclusion of PE sure to segmental level, uncertainties beyond; 4 = good-quality images, exclusion of PE to the subsegmental level surely possible; 5 = excellent to the very peripheral branches) and on a diagnostic Picture 3

Academic Radiology, Vol ■, No ■■, ■■ 2016

ZHAO ET AL

Archiving and Communication Systems (PACS) workstation with the same brightness and resolution on the same viewing monitor over a period of 2 weeks. All the images were randomized, and the readers were blinded to the NI settings. Discordance in scores between the readers was resolved by having the readers discuss and come to a consensus. Before their assessments, the readers also were instructed on the criteria for image grading, and as a group they assessed five test cases that were not included in the study to reduce interobserver variability (18). Statistical Analysis

All statistical analyses were performed with SPSS Statistics 19.0 (SPSS Inc, Chicago, IL), and the level of statistical significance (P value) was set at <.05. Quantitative variables and categorical data were expressed as mean ± standard deviation and proportions. The chi-square test was used to test for normal distribution of demographic data; the axial length, image quality scores (subjective evaluation of image quality), CTDIvol, DLP, and ED were compared using one-way analysis of variance test for all the images of CTPA. The SNR and CNR of the images of CTPA were compared using repeated measures from the data of analysis of variance, and values obtained with the lowest NI setting were used as reference standard. Interobserver variability was assessed with kappa value of concordance to measure the degree of agreement between the two radiologists for various parameters. Agreement was determined as follows: for k values less than zero, no agree-

ment; for k values of 0–0.2, slight agreement; for k values of 0.21–0.40, fair agreement; for k values of 0.41–0.60, moderate agreement; for k values of 0.61–0.80, substantial agreement; and for k values of 0.81–0.10, almost perfect agreement. RESULTS Patient Demographics

Mean age, sex distribution, weight and height of patients, and the axial length in the three groups were summarized in Table 1. There were no significant differences in age, sex distribution, weight, height, and the axial length among the 12 subgroups (all P > .05). Quantitative and Qualitative Image Analysis

There is a mean signal intensity value; mean image noise, SNR, CNR, and subjective image scores of the CTPA images in the three groups were demonstrated in Table 2 and Figure 2. There were no significant differences in mean signal intensity value (CT value), mean image noise, SNR, CNR, and subjective image scores among the four subgroups (A1, A2, A3, A4; B1, B2, B3, B4) in groups A and B (Table 2) (Figs 1 and 2) (P > .05). In group C, mean signal intensity value, SNR, CNR, and subjective image scores showed significant differences between subgroup C4 and the other subgroups (P < .01), and C4 was lower than other subgroups (Table 2)

TABLE 2. Objective Image Quality and Subjective Image Quality with ATCM Combination of Different NI Setting Mean Signal Intensity (HU) Group A (18.5 kg/m ≤ BMI < 25 kg/m ) A1 385.7 ± 132.3 A2 382.7 ± 135.5 A3 378.6 ± 123.7 A4 379.6 ± 115.3 P value all: n/s Group B (25 kg/m2 ≤ BMI < 30 kg/m2) B1 373.6 ± 123.7 B2 370.3 ± 122.3 B3 369.6 ± 127.6 B4 370.5 ± 125.8 P value all: n/s Group C (BMI ≥ 30 kg/m2) C1 364.6 ± 123.7 C2 365.3 ± 122.3 C3 364.6 ± 123.7 C4 349.5 ± 116.8 P value 4vs1: .0082 4vs2: .0076 4vs3: .0071 rest: n/s 2

Mean Image Noise (HU)

SNR

CNR

Subjective Image Scores

15.8 ± 3.2 15.6 ± 3.3 15.4 ± 3.1 15.5 ± 3.7 all: n/s

24.6 ± 2.4 24.5 ± 2.6 24.2 ± 1.8 24.3 ± 2.8 all: n/s

26.6 ± 2.6 26.2 ± 2.4 25.9 ± 3.5 26.3 ± 2.3 all: n/s

4.54 ± 0.61 4.55 ± 0.65 4.53 ± 0.62 4.52 ± 0.59 all: n/s

16.2 ± 3.5 16.1 ± 2.7 16.3 ± 3.2 16.5 ± 31 all: n/s

23.2 ± 2.3 23.3 ± 2.6 23.1 ± 2.1 22.9 ± 1.6 all: n/s

25.3 ± 2.5 24.9 ± 2.1 24.8 ± 2.2 24.7 ± 1.9 all: n/s

4.53 ± 0.58 4.49 ± 0.60 4.50 ± 0.53 4.52 ± 0.48 all: n/s

16.3 ± 3.5 16.6 ± 2.7 16.8 ± 3.3 18.9 ± 3.6 4vs1: .0065 4vs2: .0057 4vs3: .0078 rest: n/s

22.1 ± 2.3 21.6 ± 2.6 21.7 ± 2.4 18.7 ± 1.6 4vs1: .0038 4vs2: .0031 4vs3: .0052 rest: n/s

22.8 ± 2.5 22.5 ± 2.1 22.4 ± 2.4 18.9 ± 1.9 4vs1: .0043 4vs2: .0055 4vs3: .0037 rest: n/s

4.53 ± 0.58 4.49 ± 0.60 4.48 ± 0.53 4.12 ± 0.36 4vs1: .0016 4vs2: .0025 4vs3: .0021 rest: n/s

2

ATCM, automatic tube current modulation; BMI, body mass index; CNR, contrast-to-noise ratio; HU, Hounsfield unit; NI, noise index; SNR, signal-to-noise ratio.

4

ATCM, automatic tube current modulation; BMI, body mass index; CTDIvol, CT dose index of volume; DLP, dose-length product; ED, effective dose; NI, noise index.

386.69 ± 29.68 255.23 ± 21.62 223 ± 20.32 190.08 ± 19.67 46vs26: .00 40vs26: .00 36vs26: .00 46vs36: .00 rest:n/s 12.88 ± 2.28 9.16 ± 1.63 8.02 ± 1.28 6.96 ± 1.22 46vs26: .00 40vs26: .00 36vs26: .00 46vs36: .02 rest:n/s 3.83 ± 0.62 2.93 ± 0.51 2.68 ± 0.44 2.51 ± 0.23 46vs26: .00 40vs26: .00 36vs26: .00 rest:n/s 266.53 ± 20.79 231.47 ± 19.58 202.73 ± 20.76 178.21 ± 20.26 46vs26: .00 40vs26: .00 36vs26: .01 rest:n/s 8.89 ± 1.43 7.63 ± 1.51 6.45 ± 1.44 5.93 ± 1.36 46vs26: .00 40vs26: .00 36vs26: .05 rest:n/s 245.49 ± 21.68 185.33 ± 20.16 173.21 ± 19.86 160.08 ± 19.67 46vs26: .00 40vs26: .00 36vs26: .02 rest:n/s 26 36 40 46 P value

7.21 ± 1.58 6.25 ± 1.42 5.91 ± 1.31 5.56 ± 1.22 46vs26: .00 40vs26: .03 36vs26: .04 rest:n/s

3.61 ± 0.42 2.58 ± 0.31 2.49 ± 0.24 2.41 ± 0.18 46vs26: .00 40vs26: .02 36vs26: .00 rest:n/s

ED DLP CTDIvol

25 kg/m2 ≤ BMI < 30 kg/m2

ED DLP

In the last decade, different technical methods have been introduced into clinical routine to reduce radiation dose (19–22). The initial dose reduction strategies for CT focused on decreasing tube current because there is a corresponding linear reduction in radiation dose (23,24). With 3D Smart mA (xaxis, y-axis, and z-axis modulation tube current modulation) based ATCM combined with different NI settings a new technique has recently been developed for the latest scanner. In this study, we evaluated the radiation dose and image quality in ATCM by combining different NI settings with ASiR

CTDIvol

DISCUSSION

Different NI Value

Radiation dose descriptors for three groups are shown in Table 3 and Figure 2. In groups A, B, and C, there was a significant difference in CTDIvol, DLP, and ED, and the ED in subgroups A4, B4, and C3 was decreased by 33.24%, 34.47%, and 47.75% compared to subgroups A1, B1, and C1 (P = .00).

18.5 kg/m2 ≤ BMI < 25 kg/m2

Radiation Dose

TABLE 3. Radiation Dose with ATCM Combination of Different NI Setting at Different BMI

(Figs 1 and 2). The CNR, SNR, and subjective image scores in subgroup C4 were 15.4% (3.4/22.1), 17.1% (3.9/22.8), and 9.1% (0.41/4.53), respectively, lower than those values in subgroup C1 (P < .05). There was no significant difference among the other subgroups in group C (C1, C2, C3) in signal intensity value, SNR, CNR, and subjective image scores (P > .05) (Table 2) (Figs 1 and 2). There was moderate to substantial interobserver agreement between the two radiologists for subjective image quality criteria assessed in all the CTPA examinations (k = 0.61–0.8).

DLP CTDIvol

BMI ≧ 30 kg/m2 Figure 2. Graph showing the comparisons of signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and CT dose index of volume (CTDIvol) among automatic tube current modulation (ATCM) with different noise index (NI) value in different body mass index (BMI) values. NI = 46 demonstrates the worse SNR and CNR (P < .01) in BMI ≥ 31 kg/m 2 , whereas there was no statistical difference in SNR (P = .457 and .376) and CNR (P = .448 and .489) at 18.5 kg/m2 ≤ BMI < 25 kg/m2 and 25 kg/m2 ≤ BMI < 30 kg/m2. In groups A, B, and C, there was a significant difference in CTDIvol in subgroups A4, B4, and C3 and was decreased by 22.88%, 33.30%, and 37.89% compared to subgroups A1, B1, and C1 (P = .000).

6.22 ± 1.36 3.96 ± 0.48 3.25 ± 0.39 2.54 ± 0.33 46vs26: .00 40vs26: .00 36vs26: .00 46vs36: .03 rest:n/s

CTPA USING AUTOMATIC TUBE CURRENT MODULATION

ED

Academic Radiology, Vol ■, No ■■, ■■ 2016

5

ZHAO ET AL

technique. 3D Smart mA-based ATCM technique is an effective way to reduce the radiation dose. The NI is a parameter with ATCM technique that adjusts mA and corresponds to the relative noise in the images. A higher NI value means the images will contain more noise and will be obtained with a lower mA (kV is not altered) and therefore a lower radiation dose. A lower NI value means the images will contain more x-ray photon and will be obtained with a higher mA (kV is not altered) and therefore a higher radiation dose, so we should use different NI values in the patients with different BMI values to reduce the radiation dose. However, reducing the radiation dose will inevitably increase the image noise and will degrade the image quality. The ASiR technique is a reconstruction approach that improves the image quality and has been used successfully to dramatically lower the radiation dose (25,26). It can also reduce the image noise induced by the lower tube current. Hague et al. (27) and Litmanovich et al. (28) have reported reduced radiation dose with the ASiR technique, and it was used in the chest and abdomen. Image quality is not limited to image noise. In clinical application, CNR and spatial resolution for vessels and other structures are also of interest. In this study, we also evaluated SNR and CNR values as well as subjective image quality evaluation, which included the spatial resolution aspect of pulmonary vessels. Our results indicated that using the combination of ATCM and ASiR algorithm, one could obtain similar image quality even with higher NI settings (lower radiation dose), which provided the bases for recommending protocols with higher NI settings (lower radiation doses) in future CTPA applications. Recently, some radiologists have focused on reduction in tube voltage with CTPA (16,29,30) because the radiation dose is approximately proportional to the square of the tube voltage quotient. But there is a threshold limit to which the tube voltage can be reduced with current technology and image processing, because there was a serious beam-hardening artifact from the intravascular contrast media that degraded the image quality (17,31). The chest has the highest contrast between the lung and its surround organs, and has a greater x-ray attenuation variation. So, the ATCM technique was very suitable for application in the CTPA examination. The x-ray attenuation follows an exponential function; even small changes in chest diameter will result in major differences in attenuation. To achieve appropriate image quality, tube current is decreased for regions of lower photon attenuation and increased for regions of higher photon attenuation. z-Axis ATCM is longitudinal current modulation and takes into account differences in body composition in the chest direction. Thus, higher attenuation in the shoulder region as compared to the midthorax is compensated for by increasing the tube current in this area and decreasing the tube current over the lungs. xy-Axis modulation is an angular modulation that takes into account differences between the (usually larger) coronal and (usually smaller) the sagittal diameter of the chest. Angular modulation involves variation of the tube current to equalize the photon flux to the detector as the x-ray tube rotates around the patient from the anteroposterior direction to the lateral direction. The initial 6

Academic Radiology, Vol ■, No ■■, ■■ 2016

value for the tube current-time product is selected based on the purpose of the examination, and tube current decreased from the initial value during one gantry rotation (32). Inada et al. (33) and Boos et al. (34) reported that reduction of the radiation dose with automatic exposure control system in CTA of lower extremity artery and aorta. The quest for optimized CT parameters is a subject that has advanced considerably in the past decade, particularly in clinical application. Emerging data on exposure to ionizing radiation from medical imaging procedures have raised concern among radiologists about keeping the radiation dose at minimum levels and maintaining the image quality (35,36). This is the first study to prospectively evaluate image quality and radiation dose of CTPA with ATCM and different NI settings in different BMI values. In this study, we found that combination of the higher NI value and the ASiR technique can reduce the image noise and improve the image quality of the CTPA, and the radiation dose was lower than conventional NI value. The averages of ED in subgroups A1, A2, A3, A4 were, respectively 3.61 ± 0.42 mSv, 2.58 ± 0.31 mSv, 2.49 ± 0.24 mSv, and 2.41 ± 0.18 mSv, and in subgroups B1, B2, B3, B4 were, respectively 3.83 ± 0.62 mSv, 2.93 ± 0.51 mSv, 2.68 ± 0.44 mSv, and 2.51 ± 0.23 mSv. Our results demonstrated that the NI value set (when NI = 46) combining the ASiR technique allows an average of 33.24% and 34.47% radiation dose reduction compared to the NI = 26 for CTPA in 18.5 kg/m2 ≤ BMI < 30 kg/m2. The averages of ED in subgroups C1, C2, C3, and C4 were, respectively 6.22 ± 1.36 mSv, 3.96 ± 0.48 mSv, 3.25 ± 0.39 mSv, and 2.54 ± 0.33 mSv. The radiation dose of NI = 40 also was reduced by 47.75% compared to the NI = 26 in BMI ≥ 30 kg/m2. But there was no significant difference in mean image signal (mean image signal of pulmonary trunk, main left and right pulmonary arteries, left and right pulmonary lobar artery was 376.5 ± 123.7) and mean image noise (mean image noise of pulmonary trunk, main left and right pulmonary arteries, left and right pulmonary lobar artery was 15.8 ± 3.1) between NI = 26 and 46 in 18.5 kg/m2 ≤ BMI < 30 kg/m2, and two radiologists also thought that the overall image quality in the NI = 46 and 40 was similar to the NI = 26 in groups A, B, and C. Body weight and BMI are directly proportional to image noise and inversely related to arterial enhancement (37,38). As body weight increases, image noise increases and arterial enhancement decreases, a phenomenon that becomes more evident in the peripheral vasculature. Our results showed that when using the same NI value, the mean image noise in the higher BMI group was higher than that in the lower BMI group (18.9 ± 3.6 HU vs. 16.2 ± 3.1 HU). In addition, the 3rd and 4th branches of the pulmonary display were not as clearly in the higher BMI group as in the lower BMI group. As a result, the optimal NI value in the higher BMI was lower than it was in the lower BMI with the CTPA. The optimal NI value was 46 in 18.5 kg/m2 ≤ BMI < 25 kg/m2 and 25 kg/m2 ≤ BMI < 30 kg/m2, but the optimal NI was 40 in BMI > 30 kg/m2. The lower NI setting would ensure a clinically acceptable image quality in patients with the higher BMI. So

Academic Radiology, Vol ■, No ■■, ■■ 2016

the NI value in the CTPA with BMI ≥ 30 kg/m2 was lower than it was with 18.5 kg/m2 ≤ BMI < 25 kg/m2. Therefore, a higher NI value can be used with a smaller BMI (BMI < 30 kg/m2) to further reduce the radiation dose in CTPA and maintain image quality. There are several limitations in our study. First, this study included a small sample size. Second, organ dose modulation (ODM) is a new tube current modulation technique that provides additional organ-specific dose modulation. However, the technique was not equipped on our CT scanner. Third, patient scans in all the groups were not acquired consecutively, but there were no significant differences in patient characteristics (age, height, weight, and gender distribution) in all the groups. In conclusion, patient BMI-dependent NI should be used clinically, and NI settings that are higher than the recommended value may be used in CTPA with ATCM and ASiR to maximize radiation dose reduction while maintaining diagnostic image quality. The recommended NI value is 46 in CTPA examination at 18.5 kg/m2 ≤ BMI < 25 kg/m2 and 25 kg/m2 ≤ BMI < 30 kg/m2. The recommended NI value is 40 in CTPA examination at BMI ≥ 30 kg/m2. REFERENCES 1. Kuriakose J, Patel S. Acute pulmonary embolism. Radiol Clin North Am 2010; 48:31–50. 2. Heyer CM, Mohr PS, Lemburg SP, et al. Image quality and radiation exposure at pulmonary CT angiography with 100- or 120 kVp protocol: prospective randomized study. Radiology 2007; 245:577–583. 3. Winer-Muram HT, Rydberg J, Johnson MS, et al. Suspected acute pulmonary embolism: evaluation with multi-detector row CT versus digital subtraction pulmonary angiography. Radiology 2004; 233:806–815. 4. Wiest PW, Locken JA, Heintz PH, et al. CT scanning: a major source of radiation exposure. Semin Ultrasound CT MR 2002; 23:402–410. 5. Pontana F, Pagniez J, Duhamel A, et al. Reduced-dose low-voltage chest CT angiography with sinogram-affirmed iterative reconstruction versus standard-dose filtered back projection. Radiology 2013; 267:609–618. 6. Singh S, Kalra MK, Shenoy-Bhangle AS, et al. Radiation dose reduction with hybrid iterative reconstruction for pediatric CT. Radiology 2012; 263:537–546. 7. Pontana F, Duhamel A, Pagniez J, et al. Chest computed tomography using iterative reconstruction vs filtered back projection. II. Image quality of low-dose CT examinations in 80 patients. Eur Radiol 2011; 21:636– 643. 8. Baumueller S, Winklehner A, Karlo C, et al. Low-dose CT of the lung: potential value of iterative reconstructions. Eur Radiol 2012; 22:2597– 2606. 9. Jia Y, Yue D, Ning H, et al. Low-dose computed tomography with adaptive statistical iterative reconstruction and low tube voltage in craniocervical computed tomographic angiography: impact of body mass index. J Comput Assist Tomogr 2015; 39:774–780. 10. Prakash P, Kalra MK, Kambadakone AK, et al. Reducing abdominal CT radiation dose with adaptive statistical iterative reconstruction technique. Invest Radiol 2010; 45:201–210. 11. Kallen JA, Coughlin BF, O’Loughlin MT, et al. Reduced Z-axis coverage multidetector CT angiography for suspected acute pulmonary embolism could decrease dose and maintain diagnostic accuracy. Emerg Radiol 2010; 17:31–35. 12. Bongartz G, Golding SJ, Jurik AJ. European guidelines for multi slice computed tomography: report EUR 16262 EN 2004. Luxembourg: European Commission, 2004. 13. Viteri-Ramírez G, García-Lallana A, Sim-on-Yarza I, et al. Low radiation and low-contrast dose pulmonary CT angiography: comparison of 80 kVp/60 ml and 100 kVp/80 ml protocols. Clin Radiol 2012; 67:833– 839.

CTPA USING AUTOMATIC TUBE CURRENT MODULATION

14. Spielmann AL, Nelson RC, Lowry CR, et al. Liver: single breath-hold dynamic subtraction CT with multi-detector row helical technology feasibility study. Radiology 2002; 222:278–283. 15. Yuan R, Shuman WP, Earls JP, et al. Reduced iodine load at CT pulmonary angiography with dual-energy monochromatic imaging: comparison with standard CT pulmonary angiography – a prospective randomized trial. Radiology 2012; 262:290–297. 16. Schueller-Weidekamm C, Schaefer-Prokop CM, Weber M, et al. CT angiography of pulmonary arteries to detect pulmonary embolism: improvement of vascular enhancement with low kilovoltage settings. Radiology 2006; 241:899–907. 17. Fanous R, Kashani H, Jimenez L, et al. Image quality and radiation dose of pulmonary CT angiography performed using 100 and 120 kVp. AJR Am J Roentgenol 2012; 199:990–996. 18. Behrendt FF, Schmidt B, Plumhans C, et al. Image fusion in dual energy computed tomography: effect on contrast enhance signal-to-noise ratio and image quality in computed angiography. Invest Radiol 2009; 44:l– 6. 19. Vollmar SV, Kalender WA. Reduction of dose to the female breast in thoracic CT: a comparison of standard-protocol, bismuth-shielded, partial and tube-current-modulated CT examinations. Eur Radiol 2008; 18:1674– 1682. 20. Khawaja RD, Singh S, Blake M, et al. Ultralow-dose abdominal computed tomography: comparison of 2 iterative reconstruction techniques in a prospective clinical study. J Comput Assist Tomogr 2015; 39:489– 498. 21. Meer AB, Basu PA, Baker LC, et al. Exposure to ionizing radiation and estimate of secondary cancers in the era of high-speed CT scanning: projections from the Medicare population. J Am Coll Radiol 2012; 9:245– 250. 22. Lambert JW, Phelps AS, Courtier JL, et al. Image quality and dose optimisation for infant CT using a paediatric phantom. Eur Radiol 2015; 26:1–9. 23. Kubo T, Lin PJ, Stiller W, et al. Radiation dose reduction in chest CT: a review. AJR Am J Roentgenol 2008; 190:335–343. 24. Tack D, Maertelaer V, Petit W, et al. Multi-detector row CT pulmonary angiography: comparison of standard-dose and simulated low-dose techniques. Radiology 2005; 236:318–325. 25. Lee SH, Kim MJ, Yoon CS, et al. Radiation dose reduction with the adaptive statistical iterative reconstruction (ASIR) technique for chest CT in children: an intra-individual comparison. Eur J Radiol 2012; 81:e938– e943. 26. Vardhanabhuti V, Loader RJ, Mitchell GR, et al. Image quality assessment of standard- and low-dose chest CT using filtered back projection, adaptive statistical iterative reconstruction, and novel model-based iterative reconstruction algorithms. AJR Am J Roentgenol 2013; 200:545– 552. 27. Hague CJ, Krowchuk N, Alhassan D, et al. Qualitative and quantitative assessment of smoking-related lung disease: effect of iterative reconstruction on low-dose computed tomographic examinations. J Thorac Imaging 2014; 29:350–356. 28. Litmanovich DE, Tack DM, Shahrzad M, et al. Dose reduction in cardiothoracic CT: review of currently available methods. Radiographics 2014; 34:1469–1489. 29. Heyer CM, Mohr PS, Lemburg SP, et al. Image quality and radiation exposure at pulmonary CT angiography with 100- or 120-kVp protocol: prospective randomized study. Radiology 2007; 245:577–583. 30. Matsuoka S, Hunsaker A, Gill PR, et al. Vascular enhancement and image quality of MDCT pulmonary angiography in 400 cases: comparison of standard and low kilovoltage settings. AJR Am J Roentgenol 2009; 192:1651–1656. 31. Szucs-Farkas Z, Kurmann L, Strautz T, et al. Patient exposure and image quality of low-dose pulmonary computer tomography angiography: comparison of 100- and 80-kVp protocols. Invest Radiol 2008; 43:871–876. 32. Brenner DJ, Doll R, Goodhead DT, et al. Cancer risks attributable to low doses of ionizing radiation: assessing what we really know. Proc Natl Acad Sci USA 2003; 100:13761e6. 33. Inada S, Masuda T, Maruyama N, et al. Study of CT automatic exposure control system (CT-AEC) optimization in CT angiography of lower extremity artery by considering contrast-to-noise ratio. J Med Imaging Radiat Oncol 2016; 72:21–30. 34. Boos J, Aissa J, Lanzman RS, et al. CT angiography of the aorta using 80 kVp in combination with sinogram-affirmed iterative reconstruction and automated tube current modulation: effects on image quality and

7

ZHAO ET AL

radiation dose. J Med Imaging Radiat Oncol 2016; 62:187–193. doi:10.1111/1754-9485. 35. Brenner DJ, Hall EJ. Computed tomography – an increasing source of radiation exposure. N Engl J Med 2007; 357:2277e84. 36. Fazel R, Krumholz HM, Wang Y, et al. Exposure to low-dose ionizing radiation from medical imaging procedures. N Engl J Med 2009; 361:849e57.

8

Academic Radiology, Vol ■, No ■■, ■■ 2016

37. Bae KT, Seeck BA, Hildebolt CF, et al. Contrast enhancement in cardiovascular MDCT: effect of body weight, height, body surface area, body mass index, and obesity. AJR Am J Roentgenol 2008; 190:777e84. 38. Bae KT, Tao C, Gurel S, et al. Effect of patient weight and scanning duration on contrast enhancement during pulmonary multidetector CT angiography. Radiology 2007; 242:582e.