- Email: [email protected]
Low dose dual-source CT angiography in infants with complex congenital heart disease: A randomized study
European Journal of Radiology 81 (2012) e789–e795
Contents lists available at SciVerse ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Low dose dual-source CT angiography in infants with complex congenital heart disease: A randomized study Yang Gao a , Bin Lu a,∗ , Zhihui Hou a , Fangfang Yu a , Huili Cao a , Lei Han a , Runze Wu b a Department of Radiology, Cardiovascular Institute and Fu Wai Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, #167 Bei-Li-Shi Street, Beijing 100037, China b Siemens Healthcare, #7 Wang Jing Zhong Huan Nan Lu, Beijing 100102, China
a r t i c l e
i n f o
Article history: Received 28 November 2011 Received in revised form 18 March 2012 Accepted 19 March 2012 Keywords: Complex congenital heart disease Prospective ECG-triggering Dual-source CT angiography Radiation dose Image quality
a b s t r a c t Objectives: To investigate the radiation dose and image quality of prospective ECG-triggering dual-source CT angiography in infants with complex congenital heart disease (CHD) in comparison with retrospective ECG-gated scanning. Methods: Ninety-six infants less than 1 year old (60/36 male/female, age: 4.8 ± 2.7 months, weight: 5.8 ± 1.8 kg) with complex CHD were enrolled. Three image acquisition protocols were set: group 1: 80 kV, 100 mA, retrospective ECG-gated protocol; group 2: 80 kV, 100 mA, prospective ECG-triggering protocol with acquisition window of 380 ms; group 3: 80 kV, 100 mA, prospective ECG-triggering protocol with acquisition window of 200 ms. Patients were selected to any one of the protocols randomly. The signal-to-noise ratios (SNR) were calculated in the ascending aorta and the pulmonary artery trunk. Image quality was assessed by a five-point score. A score of <3 represents non-diagnostic. Effective radiation dose (ED) was calculated. Results: Image quality score of groups 1, 2 and 3 were 4.1 ± 0.4, 4.0 ± 0.6 and 4.2 ± 0.6 (p = 0.224). SNR of ascending aorta and pulmonary artery trunk among them had no statistical difference (all p > 0.05). The average ED (median) of groups 1, 2 and 3 were 1.17 ± 0.07 mSv (1.25 mSv), 0.72 ± 0.24 mSv (0.78 mSv) and 0.48 ± 0.41 mSv (0.39 mSv). Any two of the three groups had significant differences (all p < 0.001). Conclusion: Prospective ECG-triggering DSCT angiography associated with a significantly lower ED than retrospective protocol, while maintaining image quality for diagnosis. Prospective ECG-triggering DSCT angiography could be used as a very important second-line diagnostic tool in infants with complex CHD. © 2012 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Echocardiography is the first-line choice for pediatrics with congenital heart disease (CHD). It is subject to operator experience and limited by acoustic windows which restricted its evaluation of the coronary arteries and extra-cardiac vascular structures. Cardiac catheterization is always considered the gold standard for measuring pressure of cardiac chamber and confirming diagnosis of CHD. But it is an invasive procedure with an inherent risk of catheter complications such as vessel damage, bleeding and infection. The potential procedure-related mortality is around 1% [1]. Magnetic resonance imaging (MRI) is not associated with the potentially
∗ Corresponding author. Tel.: +86 10 88398052; fax: +86 10 68313012. E-mail addresses: [email protected] (Y. Gao), [email protected] (B. Lu), [email protected] (Z. Hou), [email protected] (F. Yu), [email protected] (H. Cao), hanlei [email protected] (L. Han), [email protected] (R. Wu). 0720-048X/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejrad.2012.03.023
harmful effects of ionizing radiation exposure and can provide anatomic and functional assessment of the heart. It has proven to be useful in pediatric patients. However MRI usually consists of multiple scans with a long time, it is limited in seriously ill or uncooperative patients. With faster volume coverage and higher temporal resolution, dual-source CT (DSCT) cardiac imaging has been used more frequently in complex CHD pediatrics with high resting heart rate in order to make accurate diagnosis and precise anatomic information for planning therapeutic procedure. Because of the potentially harmful effects of ionizing radiation, CT use in pediatrics must be determined on an individual basis by weighing the potential risks and benefits for each patient. Retrospective ECG-gated scan with lower kV and automatic tube current modulation could reduce the effective dose. Recently, prospective ECG-triggering techniques has been shown to decrease radiation exposure in cardiac CT [2]. With this technique, the X-ray tube is turned on only at predefined time points in the cardiac cycle. Therefore the radiation dose reduced significantly than retrospective scanning [3,4]. There have
e790
Y. Gao et al. / European Journal of Radiology 81 (2012) e789–e795
Table 1 Diagnosis of CHD in infants with different scanning protocols.
2.2. DSCT data acquisition
Diagnosis of CHD
Retrospective ECG-gated data acquisition (n = 32)
Prospective ECG-triggering data acquisition (n = 64)
COA with ASD, VSD, PDA COA with ECD, DORV, TGA IAA DORV APVC Tetralogy of Fallot with anomalous origin of coronary artery Pulmonary atresia Tricuspid atresia DORV with ECD, TGA Absence of pulmonary artery Anomalous origin of pulmonary artery Pulmonary vein stenosis Post-operation of B–T shunt or Gleen shunt Atrial isomerism with hepatic segment of inferior vena cava absence
8 2 2 1 6 1
16 6 4 2 10 2
3 2 2 1 1 1 1
7 2 4 2 2 2 3
1
2
COA, coarctation of the aorta; ASD, atrial septal defect; VSD, ventricular septal defect; PDA, patent ductus arteriosus; ECD, endocardial cushion defect; DORV, double outlet right ventricle; TGA, transposition of the great arteries; IAA, interrupted aortic arch; APVC, anomalous pulmonary venous connection.
been few randomized studies demonstrating the difference of radiation dose between different scanning procedures in infants. The aim of this randomized study was to assess the radiation dose and image quality of prospective ECG-triggering scanning with different acquisition time compared with retrospective ECG-gated tube current modulation in infants with complex CHD. 2. Methods 2.1. Study design and patients A total of 96 consecutive pediatric patients (60 male, 36 female) less than 1 year old (mean age 4.8 ± 2.7 months, 21 days to 1 year) with complex congenital heart disease previously assessed by echocardiography were examined using DSCT between March 2010 and February 2011. Weight range was 2.9–12.5 kg. Complex congenital heart disease was defined as congenital heart disease with more than one separate cardiovascular anomaly. The investigations were performed to answer specific anatomical questions raised by inconclusive echocardiography or angiography before operation or for postoperative evaluation (Table 1). The possible adverse effects of contrast medium injection and radiation exposure were explained to their legal guardians. They all signed informed consent forms. Study procedures were in accordance with the Declaration of Helsinki and the guidelines of the local ethics committee. We designed three different image acquisition protocols (Table 2). Group 1 was traditional retrospective ECG-gated protocol. Tube current was modulated at full dose during 35–75% of the R–R interval and at 4% of the full dose for the rest of R–R interval. The other two groups were prospective ECG-triggering protocols with acquisition time of 380 ms (group 2) and 200 ms (group 3), respectively. In order to guarantee successful scanning, we established the group of acquisition time with 380 ms. In order to evaluate the image quality when further minimizing radiation dose, we established the group of acquisition time with 200 ms. The center of the triggering window was set at 45% of the R–R interval. Tube voltage of the three protocols are 80 kV, tube current are 100 mA. A random tabulation was designed. The number of 1, 2, and 3 represented one kind of protocol respectively. Protocol was chosen to each infant at random.
All CT examinations were performed on a dual-source CT scanner (Somatom Definition, Siemens). Patients were scanned in the direction of cranio-cauda in quiet breathing after short-term sedation were performed with intravenous injection of ketamine (1 mg/kg). Iodixanol injection (320 mgI/ml, Visipaque, GE Healthcare) was applied using a single-head power injector (Stellant Dual Flow, Medrad, USA). The contrast media was injected through the antecubital vein. The volume was adjusted to the body weight: 1.5–2.0 ml/kg and a rate of 0.8–1.2 ml/s depending on the structure to be visualized. Bolus-tracking was used in a region-of-interest (ROI) at descending aorta. When attenuation threshold of ROI was more than 80 HU (Hounsfield unit), auto-delayed 6 s were triggered for the automatic scanning. A radiologist and a pediatrician were present to monitor vital signs including ECG and blood oxygen saturation during the examination. The following acquisition parameters were used: detector collimation, 32 mm × 0.6 mm × 2 mm; gantry rotation time, 0.33 s; slice thickness, 0.6 mm; field of view 200 mm × 200 mm. Tube voltage and tube current were set as follows: group 1: 80 kV, 100 mA; group 2: 80 kV, 100 mA, data acquisition window 380 ms; group 3: 80 kV, 100 mA, data acquisition window 200 ms. 2.3. Image post-processing and quality analysis All images were transferred to an external workstation (MMWP, Siemens) for data analysis. Images were reconstructed using a section thickness of 0.75 mm and a medium smooth-tissue convolution kernel (B26f). The reconstruction interval was 0.5 mm. Curved planar reformatting (CPR), maximum intensity projection (MIP), multiplanar reformatted (MPR) and volume rendering (VR) were created to visualize cardiac abnormalities depending on target structure and purpose. CPR was used to evaluate curved structures such as the pulmonary artery system. MIP was used mainly for evaluation of the cardiovascular structures. VR was used to evaluate the extracardiac structures [5]. Image evaluation was performed with a standardized window level of 100 HU and window width of 700 HU. Each subject was analyzed independently by two experienced cardiovascular radiologists. Both observers were blinded to the scanning parameters and patient characteristics such as weight, age and sex. Each data set was assessed for image noise and graded for image quality. The image noise and signal–noise ratio (SNR) were determined on 0.75-mm thick transverse slice by measuring the average density of contrast media and standard deviation in Hounsfield units within two regions of interest (ROI) (>100 pixels) in the pulmonary artery trunk and the middle of the ascending aorta at the same level. Each ROI was measured three times and the average value was calculated. Criteria for image quality were the subjective perception of image noise, SNR, soft-tissue contrast, sharpness of tissue interfaces, conspicuity of anatomic detail and degree of image degradation by streak or beam-hardening artifacts. For all patients, CT image quality was evaluated using a five-grade scoring system: 5, excellent anatomical clarity, excellent image quality; 4, good anatomical clarity, all structures clearly interpretable; 3, fair anatomical clarity, the anatomical relationships could be defined with confidence; 2, poor image quality of anatomical detail, incomplete demonstration of anatomical structures; 1, no useful information obtained. Examinations graded 3, 4 or 5 were considered sufficient for complete diagnosis [6]. Meanwhile, the reviewers assessed each coronary artery for image quality by using a reset four-point grading scale according to the criteria of adult [7]. The origin and traveling of each coronary could be evaluated clearly were given a score of 4; a score of 3
Y. Gao et al. / European Journal of Radiology 81 (2012) e789–e795
e791
Table 2 Patient characteristics and scanning protocols. Groups
N (male)
Group 1 Group 2 Group 3
32 (22) 31 (16) 33 (22)
In total
96 (60)
ECG-triggering methods
kV
mA
Acquisition window (ms)
Age (months)
Height (cm)
Weight (kg)
Heart rate (bpm)
Retrospective Prospective Prospective
80 80 80
100 100 100
– 380 200
5.3 ± 3.0 5.0 ± 2.4 4.0 ± 2.5
63.0 ± 6.7 62.8 ± 6.2 61.2 ± 7.2
6.0 ± 1.9 5.9 ± 1.6 5.6 ± 1.9
153.8 ± 16.3 155.1 ± 17.1 156.3 ± 17.0
4.8 ± 2.7
62.3 ± 6.7
5.8 ± 1.8
155.1 ± 16.6
Table 3 SNR of ascending aorta and pulmonary artery trunk of each group. Groups
SNR of aorta
SNR of pulmonary artery
Min
25%
50%
75%
Max
Min
25%
50%
75%
Max
Group 1 Group 2 Group 3
15 11 16
19 20 18
27 25 30
44 33 44
71 57 55
10 2 10
21 19 23
25 28 30
42 35 40
68 86 83
In total
11
20
27
41
71
2
21
29
38
86
SNR, signal-to-noise ratio.
corresponded to minor motion artifacts, clear origin of each vessel, unclear coursing of main branch but still can be evaluated; a score of 2, some motion artifacts, only the origin of coronary artery was visible while every main branch traveling could not be evaluated; and a score of 1, each vessel could not be evaluated included the origin of coronary artery. A score of 3 or 4 was considered acceptable and evaluable in terms of image quality for routine clinical diagnosis, while a score of 2 or 1 could not be evaluated or diagnosed. 2.4. Estimation of radiation dose The information of radiation dose was obtained from the CT system. The parameters included the volume CT dose index (CTDIvol ) and dose length product (DLP). To calculate effective dose (ED), the DLP was multiplied by a conversion coefficient k = 0.039 mSv/[mGy cm] for infants less than 1 year old [8]. 2.5. Statistical analysis The statistical analysis was performed using commercially available software SPSS 17.0 (SPSS, Inc. Chicago, IL) and a p-value of <0.05 was considered as significant. Quantitative data were expressed as means ± standard deviations. Image quality score of the three groups and the value of SNR, CTDI, DLP and ED were non-normal distribution (Kruskal–Wallis H test), which were compared using Mann–Whitney U test among each groups. Inter-observer agreement was tested by Kappa value. 3. Results 3.1. Patient characteristics The final diagnosis of the 96 patients with complex CHD were coarctation of the aorta with other abnormalities (n = 32), interrupted aortic arch (n = 6), double outlet right ventricle (n = 3),
anomalous pulmonary venous connection (n = 16), tetralogy of Fallot with anomalous origin of coronary artery (n = 3), pulmonary atresia (n = 10), tricuspid atresia (n = 4), double outlet right ventricle with endocardial cushion defect or transposition of the great arteries (n = 6), absence of pulmonary artery (n = 3), anomalous origin of pulmonary artery (n = 3), pulmonary vein stenosis (n = 3), post-operation of B–T shunt or Gleen shunt (n = 4), atrial isomerism with hepatic segment of inferior vena cava absence (n = 3). There are 32 pediatrics including 22 boys performed retrospective ECGgated protocol (group 1), 31 pediatrics including 16 boys performed prospective ECG-triggering protocol with acquisition window of 380 ms (group 2) and 32 pediatrics including 22 boys performed prospective ECG-triggering protocol with acquisition window of 200 ms (group 3). Characteristics of age, height, weight and heart rate during scanning among the three groups have no significant differences (Table 2). There were four cases (4.17%) had congenital malformations of coronary artery. One was LCX originated from the right-coronary sinus. One was RCA originated from the left coronary sinus. One was LAD originated from pulmonary artery. The other was LAD and RCA both originated from the non-coronary sinus.
3.2. Subjective evaluation of image quality Interobserver agreement of overall image quality was reached in 93 (96.9%) studies (Kappa = 0.911). Disagreement occurred in 3 studies, all of them with only one point of difference. The average subjective image quality score of all the patients was 4.1 ± 0.5 (range 3–5). Diagnostic images (images graded 3 or more) were obtained in all of the consecutive examinations (100%). All of the image quality was well enough to meet diagnosis. The average image quality score of each group was 4.1 ± 0.4 (range 3–5), 4.0 ± 0.6 (range 3–5) and 4.2 ± 0.6 (range 3–5), respectively. There was no significant difference among them (p = 0.341). The mean
Table 4 Radiation dose of each group. Groups
Scanning slices
25%
50%
75%
25%
50%
75%
25%
50%
75%
Group 1 Group 2 Group 3
226 ± 47 223 ± 39 205 ± 41
2.06 1.41 0.80
2.20 1.56 0.98
2.32 1.67 1.65
27.00 14.00 7.50
32.00 20.00 10.00
35.00 23.00 20.00
1.05 0.55 0.29
1.25 0.78 0.39
1.37 0.90 0.78
In total
218 ± 43
1.27
1.66
2.07
12.00
20.00
27.75
0.47
0.78
0.98
CTDI (mGy)
DLP (mGy/cm)
ED (mSv)
e792
Y. Gao et al. / European Journal of Radiology 81 (2012) e789–e795
image quality score of all the coronary arteries was 3.5 ± 1.0. Image quality scores were considered 3 and 4 in 80/96 (83.33%) of patients. 3.3. Objective evaluation of image quality The average attenuation in both the ascending aorta and pulmonary trunk was 459 ± 111 HU and 452 ± 116 HU. The median noise in the ascending aorta and pulmonary trunk was 17 HU and 19 HU. The median signal-to-noise ratio (SNR) in the ascending aorta and pulmonary trunk was 27 and 29. SNR of ascending aorta and pulmonary artery in all groups had no statistical differences (all p > 0.05) (Table 3). 3.4. Radiation dose estimation The radiation dose parameters were summarized in Table 4. The average DLP and ED of group 1, group 2 and group 3 were 30.11 mGy/cm, 18.52 mGy/cm, 12.29 mGy/cm and 1.17 mSv, 0.72 mSv and 0.48 mSv respectively. The value of CTDI, DLP and ED had significantly statistical differences among any two of the three groups (all p < 0.001). Radiation dose using prospective ECG-triggering protocol was lower than retrospective ECG-gated protocol. Narrowed data acquisition window led to lower radiation
Fig. 1. Effective dose of the three groups. (Any two of them have significant difference, all p < 0.001.)
Fig. 2. A 3-month-old boy with the diagnosis of coarctation of the aorta (COA) and ventricular septal defect (VSD). Prospective ECG-triggering protocol with acquisition window of 380 ms was performed. Effective dose was 0.68 mSv. (a) Axial image shows ventricular septal defect (VSD) below aortic valve. (b) Axial image and (c) volume rendered image showed coarctation of the upper descending aorta (arrow). AA, ascending aorta; LA, left atrium; RV, right ventricle; PA, pulmonary artery; DA, descending aorta.
Y. Gao et al. / European Journal of Radiology 81 (2012) e789–e795
e793
Fig. 3. A 11-month-old girl with the diagnosis of Tetralogy of Fallot. Retrospective ECG-gated protocol of DSCT was performed after B–T shunt operation on this patient. Effective dose was 0.85 mSv. (a) Axial image showed ventricular septal defect (VSD) and the aorta ride between the left and right ventricles. (b) Volume rendered image and (c) multiplanar reformatted images showed a clear shunt vessel connected between right carotid artery and right pulmonary artery (arrow). AO, aorta; LV, left ventricle; RV, right ventricle; DA, descending aorta; RCCA, right common carotid artery; RPA, right pulmonary artery; ARSA, aberrant right subclavian artery; LA, left atrium; LV, left ventricle.
dose (Fig. 1). Figs. 2–4 show the images of infants using different scanning protocols.
4. Discussion The purpose of CT exams on infants with complex CHD was to make clear of the diagnosis before surgery or evaluate the operation efficacy to prepare for the following therapy. DSCT scanners with adequate temporal and spatial resolution can be used for demonstrating cardiac and extra-cardiac anatomic malformations in comparison with other imaging techniques. Pediatric cardiac CT is challenging because of the small body size, unstable vital signs. How to reduce the radiation dose on CT scans is the most important issue for its clinical use, especially important for young patients. We must choose optimal protocol to decrease radiation dose as low as reasonably achievable (ALARA), while meet demand for the diagnostic task. Most of previous studies analyzed the radiation dose and image quality retrospectively. Our study is a prospective randomized design. Infants with complex CHD less than 1 year old were selected to use retrospective or prospective scanning protocol randomly regardless of his or her size or weight.
4.1. Radiation exposure Minimization of the radiation exposure delivered by CT is an important issue particularly for pediatrics. Lowering tube voltage or current is the direct way of achieving radiation dose reduction. For pediatric patients, due to less attenuation in the body, the noise level does not increase significantly with the decrease of kV for the same radiation dose. Therefore, for iodine contrast enhanced exams, a lower kV can be used to improve the contrast enhancement without increasing the noise [9]. Protocols using 80 kV increase sensitivity to contrast medium by 50% compared to 120 kV acquisitions [10]. Low kV techniques favors safety issues on both sides – contrast medium and radiation exposure and have been widely used in pediatric CT to reduce radiation exposure without impairing image quality. In the study by Lee et al., the average DLP in neonates with CHD was 47 mGy/cm at 80 kV and 162 mGy/cm at 120 kV [11]. In our study the average DLP of all patients was 20 mGy/cm at 80 kV. Due to excessively high heart rate on infant patients, the retrospective ECG-gated scanning usually performed to acquire higher chance for good image quality. Tube current is set to prescribed value (100%) at specific phase of R–R interval and reduced to 4–20%
e794
Y. Gao et al. / European Journal of Radiology 81 (2012) e789–e795
Fig. 4. A 1-month-old boy with the diagnosis of total anomaly of pulmonary vein connection (TAPVC). Prospective ECG-triggering protocol with acquisition window of 200 ms was performed. Effective dose was 0.58 mSv. (a) Axial image showed atrial septal defect (ASD) and the right ventricle enlarged. (b) Coronal MIP and (c) volume rendered image showed the left and right pulmonary vein merged into a trunk and traveling down to the portal vein (arrow). LV, left ventricle; RV, right ventricle; PVT, pulmonary vein trunk.
at other phases. However, retrospective ECG-gated scan was associated with high radiation dose even with low tube voltage. In our study with retrospective ECG-gated spiral scan, 80 kV was used in infants less than 1 year old, tube current was modulated at full dose (100 mA) during 35–75% of the R–R interval and at 4% of the full dose for the rest of R–R interval. The average effective dose was 1.1 mSv which was significantly lower than previous study [6,12]. Recently prospectively ECG-triggering sequential data acquisition (so called “Step- and -Shoot” mode) was introduced for cardiac CTA in adults and has been employed for imaging of pediatric patients with CHD in order to reduce radiation dose [13]. Goo et al. found a low dose ECG-synchronization technique that can reduce the CT dose to 0.2–0.7 mSv in newborns and infants [12]. Cheng et al. reported the use of prospective ECG-triggering sequential CTA in pediatric patients with a mean ED of 0.38 mSv in a cohort of 35 children up to the age of 6 years [14]. To account for the higher heart rates generally observed in children the data acquisition window was centered at 40% in the end-systole [3]. In our study, prospective ECG-triggering was used with the data acquisition window of 380 ms and 200 ms centered at 45% in the cardiac cycle, tube voltage was set at 80 kV. The average ED was 0.72 mSv and 0.48 mSv for scans with 380 ms and 200 ms acquisition window. The shorter
the data acquisition window is, the lower the radiation exposure is while image quality was not affected. For diagnostic cardiac catheterizations, a median effective dose of 4.6 mSv was found for pediatric patients [15]. Therapeutic procedures resulted in a higher median effective dose of 6.0 mSv because of the prolonged use of fluoroscopy and the larger number of cine runs [15]. A study of 4952 consecutive pediatric catheterization procedures found that one or more complications occurred in 436 patients (8.8%) [1]. From our perspective, as mean radiation dose declines significantly, low-dose CTA seems better than catheter angiography in evaluating anatomical structure of children with complex CHD. 4.2. Image quality Our results indicated that using prospective ECG-triggering protocol, successfully achieved diagnostic image quality in 100% of examinations and significantly lowered radiation doses ass compared to a retrospective ECG-gated protocol. In this study, we used short-acting sedative drugs in order to minimize motion artifacts. Sleepy babies ensure good image quality with a short total examination time. We found that DSCT acquired images of diagnostic quality in all of the complex CHD patients with a very fast heart rate and with good interobserver agreement. Only minor breathing
Y. Gao et al. / European Journal of Radiology 81 (2012) e789–e795
related artifacts were observed but did not impair diagnostic image quality. For cardiovascular CT examination in infants, on the basis of 80 kV and 100 mA, using prospective ECG-triggering sequential data acquisition could reduce the radiation dose without compromising image quality. Lee et al. studied 14 neonates with CHD and found that MDCT had a diagnostic accuracy of 98%. They suggested that MDCT could replace cardiac catheterization for this indication [11]. Noninvasive CTA had shorter examination time and higher spatial resolution than magnetic resonance imaging (MRI). In comparison with cardiac catheterization, CT angiography was a safe, simple method, requiring less contrast agent and fewer complications on pediatric cardiac imaging. As the technology advances and improvements are made in scan protocols, ionizing radiation received during a CT exam will be further reduced. In our hospital almost all of the pediatric patients with complex CHD are performed with prospective ECG-triggering protocol currently. 5. Limitations Our study is a prospective randomized design for image quality and safety issues. Radiation dose has a close relation to body weight. The impact of body weight was not take into account for randomization in this study. In addition, part of infants have done conventional cardiovascular angiography. We cannot draw conclusions regarding the optimal imaging strategy in patients with complex CHD on the basis of our data. In our hospital, DSCT is used as an important second-line modality to complete or confirm diagnosis. 6. Conclusion The finding of our study shows that high spatial resolution and fast data acquisition make DSCT an obvious modality to evaluate the unstable neonate with a small heart and complex anatomy. Prospective ECG-triggering DSCT reduce radiation dose to a significantly low level while maintaining image quality. Prospective ECG-triggering DSCT angiography could be used as a very important diagnostic tool in infants with complex CHD.
e795
Conflict of interest None of the authors have any conflicting interests. References [1] Vitiello R, McCrindle BW, Nykanen D, et al. Complications associated with pediatric cardiac catheterization. Journal of the American College of Cardiology 1998;32:1433–40. [2] Jin KN, Park EA, Shin CI, et al. Retrospective versus prospective ECG-gated dual-source CT in pediatric patients with congenital heart disease: comparison of image quality and radiation dose. International Journal of Cardiovascular Imaging 2010;26:63–73. [3] Herzog C, Mulvihill DM, Nguyen SA, et al. Pediatric cardiovascular CT angiography: radiation dose reduction using automatic anatomic tube current modulation. American Journal of Radiology 2008;190:1232–40. [4] Mayo JR, Leipsic JA. Radiation dose in cardiac CT. American Journal of Radiology 2009;192:646–53. [5] Goo HW, Park IS, Ko JK, et al. CT of congenital heart disease: normal anatomy and typical pathologic conditions. RadioGraphics 2003;23:S147–65. [6] Saad MB, Rohnean A, Sigal-Cinqualbre A, et al. Evaluation of image quality and radiation dose of thoracic and coronary dual-source CT in 110 infants with congenital heart disease. Pediatric Radiology 2009;39:668–76. [7] Bamberg F, Abbaraa S, Schlett CL, et al. Predictors of image quality of coronary computed tomography in the acute care setting of patients with chest pain. European Journal of Radiology 2010;74(1):182–8. [8] Shrimpton PC. Assessment of patient dose in CT. NRPB-PE/1/2004. NRPB, Chilton. Also published as Appendix C of the 2004 CT Quality Criteria (MSCT, 2004) at: http://www.msct.eu/CT Quality Criteria.htm. [9] McCollough CH, Primak AN, Braunc N, et al. Strategies for reducing radiation dose in CT. Radiologic Clinics of North America 2009;47:27–40. [10] Sigal-Cinqualbre AB, Hennequin R, Abada HT, et al. Low-kilovoltage multidetector row chest CT in adults: feasibility and effect on image quality and iodine dose. Radiology 2004;231:169–74. [11] Lee T, Tsai IC, Fu YC, et al. Using multidetector-row CT in neonates with complex congenital heart disease to replace diagnostic cardiac catheterization for anatomical investigation: initial experiences in technical and clinical feasibility. Pediatric Radiology 2006;36:1273–82. [12] Goo HW, Seo DM, Yun TJ, et al. Coronary artery anomalies and clinically important anatomy in patients with congenital heart disease: multislice CT findings. Pediatric Radiology 2009;39:265–73. [13] Stolzmann P, Leschka S, Scheffel H, et al. Dual-source CT in step-and-shoot mode: noninvasive coronary angiography with low radiation dose. Radiology 2008;49:71–80. [14] Cheng ZP, Wang XM, Duan YH, et al. Low-dose prospective ECG-triggering dualsource CT angiography in infants and children with complex congenital heart disease: first experience. European Radiology 2010;20:2503–11. [15] Bacher K, Bogaert E, Lapere R, et al. Patient-specific dose and radiation risk estimation in pediatric cardiac catheterization. Circulation 2005;111:83–9.
Contents lists available at SciVerse ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Low dose dual-source CT angiography in infants with complex congenital heart disease: A randomized study Yang Gao a , Bin Lu a,∗ , Zhihui Hou a , Fangfang Yu a , Huili Cao a , Lei Han a , Runze Wu b a Department of Radiology, Cardiovascular Institute and Fu Wai Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, #167 Bei-Li-Shi Street, Beijing 100037, China b Siemens Healthcare, #7 Wang Jing Zhong Huan Nan Lu, Beijing 100102, China
a r t i c l e
i n f o
Article history: Received 28 November 2011 Received in revised form 18 March 2012 Accepted 19 March 2012 Keywords: Complex congenital heart disease Prospective ECG-triggering Dual-source CT angiography Radiation dose Image quality
a b s t r a c t Objectives: To investigate the radiation dose and image quality of prospective ECG-triggering dual-source CT angiography in infants with complex congenital heart disease (CHD) in comparison with retrospective ECG-gated scanning. Methods: Ninety-six infants less than 1 year old (60/36 male/female, age: 4.8 ± 2.7 months, weight: 5.8 ± 1.8 kg) with complex CHD were enrolled. Three image acquisition protocols were set: group 1: 80 kV, 100 mA, retrospective ECG-gated protocol; group 2: 80 kV, 100 mA, prospective ECG-triggering protocol with acquisition window of 380 ms; group 3: 80 kV, 100 mA, prospective ECG-triggering protocol with acquisition window of 200 ms. Patients were selected to any one of the protocols randomly. The signal-to-noise ratios (SNR) were calculated in the ascending aorta and the pulmonary artery trunk. Image quality was assessed by a five-point score. A score of <3 represents non-diagnostic. Effective radiation dose (ED) was calculated. Results: Image quality score of groups 1, 2 and 3 were 4.1 ± 0.4, 4.0 ± 0.6 and 4.2 ± 0.6 (p = 0.224). SNR of ascending aorta and pulmonary artery trunk among them had no statistical difference (all p > 0.05). The average ED (median) of groups 1, 2 and 3 were 1.17 ± 0.07 mSv (1.25 mSv), 0.72 ± 0.24 mSv (0.78 mSv) and 0.48 ± 0.41 mSv (0.39 mSv). Any two of the three groups had significant differences (all p < 0.001). Conclusion: Prospective ECG-triggering DSCT angiography associated with a significantly lower ED than retrospective protocol, while maintaining image quality for diagnosis. Prospective ECG-triggering DSCT angiography could be used as a very important second-line diagnostic tool in infants with complex CHD. © 2012 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Echocardiography is the first-line choice for pediatrics with congenital heart disease (CHD). It is subject to operator experience and limited by acoustic windows which restricted its evaluation of the coronary arteries and extra-cardiac vascular structures. Cardiac catheterization is always considered the gold standard for measuring pressure of cardiac chamber and confirming diagnosis of CHD. But it is an invasive procedure with an inherent risk of catheter complications such as vessel damage, bleeding and infection. The potential procedure-related mortality is around 1% [1]. Magnetic resonance imaging (MRI) is not associated with the potentially
∗ Corresponding author. Tel.: +86 10 88398052; fax: +86 10 68313012. E-mail addresses: [email protected] (Y. Gao), [email protected] (B. Lu), [email protected] (Z. Hou), [email protected] (F. Yu), [email protected] (H. Cao), hanlei [email protected] (L. Han), [email protected] (R. Wu). 0720-048X/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejrad.2012.03.023
harmful effects of ionizing radiation exposure and can provide anatomic and functional assessment of the heart. It has proven to be useful in pediatric patients. However MRI usually consists of multiple scans with a long time, it is limited in seriously ill or uncooperative patients. With faster volume coverage and higher temporal resolution, dual-source CT (DSCT) cardiac imaging has been used more frequently in complex CHD pediatrics with high resting heart rate in order to make accurate diagnosis and precise anatomic information for planning therapeutic procedure. Because of the potentially harmful effects of ionizing radiation, CT use in pediatrics must be determined on an individual basis by weighing the potential risks and benefits for each patient. Retrospective ECG-gated scan with lower kV and automatic tube current modulation could reduce the effective dose. Recently, prospective ECG-triggering techniques has been shown to decrease radiation exposure in cardiac CT [2]. With this technique, the X-ray tube is turned on only at predefined time points in the cardiac cycle. Therefore the radiation dose reduced significantly than retrospective scanning [3,4]. There have
e790
Y. Gao et al. / European Journal of Radiology 81 (2012) e789–e795
Table 1 Diagnosis of CHD in infants with different scanning protocols.
2.2. DSCT data acquisition
Diagnosis of CHD
Retrospective ECG-gated data acquisition (n = 32)
Prospective ECG-triggering data acquisition (n = 64)
COA with ASD, VSD, PDA COA with ECD, DORV, TGA IAA DORV APVC Tetralogy of Fallot with anomalous origin of coronary artery Pulmonary atresia Tricuspid atresia DORV with ECD, TGA Absence of pulmonary artery Anomalous origin of pulmonary artery Pulmonary vein stenosis Post-operation of B–T shunt or Gleen shunt Atrial isomerism with hepatic segment of inferior vena cava absence
8 2 2 1 6 1
16 6 4 2 10 2
3 2 2 1 1 1 1
7 2 4 2 2 2 3
1
2
COA, coarctation of the aorta; ASD, atrial septal defect; VSD, ventricular septal defect; PDA, patent ductus arteriosus; ECD, endocardial cushion defect; DORV, double outlet right ventricle; TGA, transposition of the great arteries; IAA, interrupted aortic arch; APVC, anomalous pulmonary venous connection.
been few randomized studies demonstrating the difference of radiation dose between different scanning procedures in infants. The aim of this randomized study was to assess the radiation dose and image quality of prospective ECG-triggering scanning with different acquisition time compared with retrospective ECG-gated tube current modulation in infants with complex CHD. 2. Methods 2.1. Study design and patients A total of 96 consecutive pediatric patients (60 male, 36 female) less than 1 year old (mean age 4.8 ± 2.7 months, 21 days to 1 year) with complex congenital heart disease previously assessed by echocardiography were examined using DSCT between March 2010 and February 2011. Weight range was 2.9–12.5 kg. Complex congenital heart disease was defined as congenital heart disease with more than one separate cardiovascular anomaly. The investigations were performed to answer specific anatomical questions raised by inconclusive echocardiography or angiography before operation or for postoperative evaluation (Table 1). The possible adverse effects of contrast medium injection and radiation exposure were explained to their legal guardians. They all signed informed consent forms. Study procedures were in accordance with the Declaration of Helsinki and the guidelines of the local ethics committee. We designed three different image acquisition protocols (Table 2). Group 1 was traditional retrospective ECG-gated protocol. Tube current was modulated at full dose during 35–75% of the R–R interval and at 4% of the full dose for the rest of R–R interval. The other two groups were prospective ECG-triggering protocols with acquisition time of 380 ms (group 2) and 200 ms (group 3), respectively. In order to guarantee successful scanning, we established the group of acquisition time with 380 ms. In order to evaluate the image quality when further minimizing radiation dose, we established the group of acquisition time with 200 ms. The center of the triggering window was set at 45% of the R–R interval. Tube voltage of the three protocols are 80 kV, tube current are 100 mA. A random tabulation was designed. The number of 1, 2, and 3 represented one kind of protocol respectively. Protocol was chosen to each infant at random.
All CT examinations were performed on a dual-source CT scanner (Somatom Definition, Siemens). Patients were scanned in the direction of cranio-cauda in quiet breathing after short-term sedation were performed with intravenous injection of ketamine (1 mg/kg). Iodixanol injection (320 mgI/ml, Visipaque, GE Healthcare) was applied using a single-head power injector (Stellant Dual Flow, Medrad, USA). The contrast media was injected through the antecubital vein. The volume was adjusted to the body weight: 1.5–2.0 ml/kg and a rate of 0.8–1.2 ml/s depending on the structure to be visualized. Bolus-tracking was used in a region-of-interest (ROI) at descending aorta. When attenuation threshold of ROI was more than 80 HU (Hounsfield unit), auto-delayed 6 s were triggered for the automatic scanning. A radiologist and a pediatrician were present to monitor vital signs including ECG and blood oxygen saturation during the examination. The following acquisition parameters were used: detector collimation, 32 mm × 0.6 mm × 2 mm; gantry rotation time, 0.33 s; slice thickness, 0.6 mm; field of view 200 mm × 200 mm. Tube voltage and tube current were set as follows: group 1: 80 kV, 100 mA; group 2: 80 kV, 100 mA, data acquisition window 380 ms; group 3: 80 kV, 100 mA, data acquisition window 200 ms. 2.3. Image post-processing and quality analysis All images were transferred to an external workstation (MMWP, Siemens) for data analysis. Images were reconstructed using a section thickness of 0.75 mm and a medium smooth-tissue convolution kernel (B26f). The reconstruction interval was 0.5 mm. Curved planar reformatting (CPR), maximum intensity projection (MIP), multiplanar reformatted (MPR) and volume rendering (VR) were created to visualize cardiac abnormalities depending on target structure and purpose. CPR was used to evaluate curved structures such as the pulmonary artery system. MIP was used mainly for evaluation of the cardiovascular structures. VR was used to evaluate the extracardiac structures [5]. Image evaluation was performed with a standardized window level of 100 HU and window width of 700 HU. Each subject was analyzed independently by two experienced cardiovascular radiologists. Both observers were blinded to the scanning parameters and patient characteristics such as weight, age and sex. Each data set was assessed for image noise and graded for image quality. The image noise and signal–noise ratio (SNR) were determined on 0.75-mm thick transverse slice by measuring the average density of contrast media and standard deviation in Hounsfield units within two regions of interest (ROI) (>100 pixels) in the pulmonary artery trunk and the middle of the ascending aorta at the same level. Each ROI was measured three times and the average value was calculated. Criteria for image quality were the subjective perception of image noise, SNR, soft-tissue contrast, sharpness of tissue interfaces, conspicuity of anatomic detail and degree of image degradation by streak or beam-hardening artifacts. For all patients, CT image quality was evaluated using a five-grade scoring system: 5, excellent anatomical clarity, excellent image quality; 4, good anatomical clarity, all structures clearly interpretable; 3, fair anatomical clarity, the anatomical relationships could be defined with confidence; 2, poor image quality of anatomical detail, incomplete demonstration of anatomical structures; 1, no useful information obtained. Examinations graded 3, 4 or 5 were considered sufficient for complete diagnosis [6]. Meanwhile, the reviewers assessed each coronary artery for image quality by using a reset four-point grading scale according to the criteria of adult [7]. The origin and traveling of each coronary could be evaluated clearly were given a score of 4; a score of 3
Y. Gao et al. / European Journal of Radiology 81 (2012) e789–e795
e791
Table 2 Patient characteristics and scanning protocols. Groups
N (male)
Group 1 Group 2 Group 3
32 (22) 31 (16) 33 (22)
In total
96 (60)
ECG-triggering methods
kV
mA
Acquisition window (ms)
Age (months)
Height (cm)
Weight (kg)
Heart rate (bpm)
Retrospective Prospective Prospective
80 80 80
100 100 100
– 380 200
5.3 ± 3.0 5.0 ± 2.4 4.0 ± 2.5
63.0 ± 6.7 62.8 ± 6.2 61.2 ± 7.2
6.0 ± 1.9 5.9 ± 1.6 5.6 ± 1.9
153.8 ± 16.3 155.1 ± 17.1 156.3 ± 17.0
4.8 ± 2.7
62.3 ± 6.7
5.8 ± 1.8
155.1 ± 16.6
Table 3 SNR of ascending aorta and pulmonary artery trunk of each group. Groups
SNR of aorta
SNR of pulmonary artery
Min
25%
50%
75%
Max
Min
25%
50%
75%
Max
Group 1 Group 2 Group 3
15 11 16
19 20 18
27 25 30
44 33 44
71 57 55
10 2 10
21 19 23
25 28 30
42 35 40
68 86 83
In total
11
20
27
41
71
2
21
29
38
86
SNR, signal-to-noise ratio.
corresponded to minor motion artifacts, clear origin of each vessel, unclear coursing of main branch but still can be evaluated; a score of 2, some motion artifacts, only the origin of coronary artery was visible while every main branch traveling could not be evaluated; and a score of 1, each vessel could not be evaluated included the origin of coronary artery. A score of 3 or 4 was considered acceptable and evaluable in terms of image quality for routine clinical diagnosis, while a score of 2 or 1 could not be evaluated or diagnosed. 2.4. Estimation of radiation dose The information of radiation dose was obtained from the CT system. The parameters included the volume CT dose index (CTDIvol ) and dose length product (DLP). To calculate effective dose (ED), the DLP was multiplied by a conversion coefficient k = 0.039 mSv/[mGy cm] for infants less than 1 year old [8]. 2.5. Statistical analysis The statistical analysis was performed using commercially available software SPSS 17.0 (SPSS, Inc. Chicago, IL) and a p-value of <0.05 was considered as significant. Quantitative data were expressed as means ± standard deviations. Image quality score of the three groups and the value of SNR, CTDI, DLP and ED were non-normal distribution (Kruskal–Wallis H test), which were compared using Mann–Whitney U test among each groups. Inter-observer agreement was tested by Kappa value. 3. Results 3.1. Patient characteristics The final diagnosis of the 96 patients with complex CHD were coarctation of the aorta with other abnormalities (n = 32), interrupted aortic arch (n = 6), double outlet right ventricle (n = 3),
anomalous pulmonary venous connection (n = 16), tetralogy of Fallot with anomalous origin of coronary artery (n = 3), pulmonary atresia (n = 10), tricuspid atresia (n = 4), double outlet right ventricle with endocardial cushion defect or transposition of the great arteries (n = 6), absence of pulmonary artery (n = 3), anomalous origin of pulmonary artery (n = 3), pulmonary vein stenosis (n = 3), post-operation of B–T shunt or Gleen shunt (n = 4), atrial isomerism with hepatic segment of inferior vena cava absence (n = 3). There are 32 pediatrics including 22 boys performed retrospective ECGgated protocol (group 1), 31 pediatrics including 16 boys performed prospective ECG-triggering protocol with acquisition window of 380 ms (group 2) and 32 pediatrics including 22 boys performed prospective ECG-triggering protocol with acquisition window of 200 ms (group 3). Characteristics of age, height, weight and heart rate during scanning among the three groups have no significant differences (Table 2). There were four cases (4.17%) had congenital malformations of coronary artery. One was LCX originated from the right-coronary sinus. One was RCA originated from the left coronary sinus. One was LAD originated from pulmonary artery. The other was LAD and RCA both originated from the non-coronary sinus.
3.2. Subjective evaluation of image quality Interobserver agreement of overall image quality was reached in 93 (96.9%) studies (Kappa = 0.911). Disagreement occurred in 3 studies, all of them with only one point of difference. The average subjective image quality score of all the patients was 4.1 ± 0.5 (range 3–5). Diagnostic images (images graded 3 or more) were obtained in all of the consecutive examinations (100%). All of the image quality was well enough to meet diagnosis. The average image quality score of each group was 4.1 ± 0.4 (range 3–5), 4.0 ± 0.6 (range 3–5) and 4.2 ± 0.6 (range 3–5), respectively. There was no significant difference among them (p = 0.341). The mean
Table 4 Radiation dose of each group. Groups
Scanning slices
25%
50%
75%
25%
50%
75%
25%
50%
75%
Group 1 Group 2 Group 3
226 ± 47 223 ± 39 205 ± 41
2.06 1.41 0.80
2.20 1.56 0.98
2.32 1.67 1.65
27.00 14.00 7.50
32.00 20.00 10.00
35.00 23.00 20.00
1.05 0.55 0.29
1.25 0.78 0.39
1.37 0.90 0.78
In total
218 ± 43
1.27
1.66
2.07
12.00
20.00
27.75
0.47
0.78
0.98
CTDI (mGy)
DLP (mGy/cm)
ED (mSv)
e792
Y. Gao et al. / European Journal of Radiology 81 (2012) e789–e795
image quality score of all the coronary arteries was 3.5 ± 1.0. Image quality scores were considered 3 and 4 in 80/96 (83.33%) of patients. 3.3. Objective evaluation of image quality The average attenuation in both the ascending aorta and pulmonary trunk was 459 ± 111 HU and 452 ± 116 HU. The median noise in the ascending aorta and pulmonary trunk was 17 HU and 19 HU. The median signal-to-noise ratio (SNR) in the ascending aorta and pulmonary trunk was 27 and 29. SNR of ascending aorta and pulmonary artery in all groups had no statistical differences (all p > 0.05) (Table 3). 3.4. Radiation dose estimation The radiation dose parameters were summarized in Table 4. The average DLP and ED of group 1, group 2 and group 3 were 30.11 mGy/cm, 18.52 mGy/cm, 12.29 mGy/cm and 1.17 mSv, 0.72 mSv and 0.48 mSv respectively. The value of CTDI, DLP and ED had significantly statistical differences among any two of the three groups (all p < 0.001). Radiation dose using prospective ECG-triggering protocol was lower than retrospective ECG-gated protocol. Narrowed data acquisition window led to lower radiation
Fig. 1. Effective dose of the three groups. (Any two of them have significant difference, all p < 0.001.)
Fig. 2. A 3-month-old boy with the diagnosis of coarctation of the aorta (COA) and ventricular septal defect (VSD). Prospective ECG-triggering protocol with acquisition window of 380 ms was performed. Effective dose was 0.68 mSv. (a) Axial image shows ventricular septal defect (VSD) below aortic valve. (b) Axial image and (c) volume rendered image showed coarctation of the upper descending aorta (arrow). AA, ascending aorta; LA, left atrium; RV, right ventricle; PA, pulmonary artery; DA, descending aorta.
Y. Gao et al. / European Journal of Radiology 81 (2012) e789–e795
e793
Fig. 3. A 11-month-old girl with the diagnosis of Tetralogy of Fallot. Retrospective ECG-gated protocol of DSCT was performed after B–T shunt operation on this patient. Effective dose was 0.85 mSv. (a) Axial image showed ventricular septal defect (VSD) and the aorta ride between the left and right ventricles. (b) Volume rendered image and (c) multiplanar reformatted images showed a clear shunt vessel connected between right carotid artery and right pulmonary artery (arrow). AO, aorta; LV, left ventricle; RV, right ventricle; DA, descending aorta; RCCA, right common carotid artery; RPA, right pulmonary artery; ARSA, aberrant right subclavian artery; LA, left atrium; LV, left ventricle.
dose (Fig. 1). Figs. 2–4 show the images of infants using different scanning protocols.
4. Discussion The purpose of CT exams on infants with complex CHD was to make clear of the diagnosis before surgery or evaluate the operation efficacy to prepare for the following therapy. DSCT scanners with adequate temporal and spatial resolution can be used for demonstrating cardiac and extra-cardiac anatomic malformations in comparison with other imaging techniques. Pediatric cardiac CT is challenging because of the small body size, unstable vital signs. How to reduce the radiation dose on CT scans is the most important issue for its clinical use, especially important for young patients. We must choose optimal protocol to decrease radiation dose as low as reasonably achievable (ALARA), while meet demand for the diagnostic task. Most of previous studies analyzed the radiation dose and image quality retrospectively. Our study is a prospective randomized design. Infants with complex CHD less than 1 year old were selected to use retrospective or prospective scanning protocol randomly regardless of his or her size or weight.
4.1. Radiation exposure Minimization of the radiation exposure delivered by CT is an important issue particularly for pediatrics. Lowering tube voltage or current is the direct way of achieving radiation dose reduction. For pediatric patients, due to less attenuation in the body, the noise level does not increase significantly with the decrease of kV for the same radiation dose. Therefore, for iodine contrast enhanced exams, a lower kV can be used to improve the contrast enhancement without increasing the noise [9]. Protocols using 80 kV increase sensitivity to contrast medium by 50% compared to 120 kV acquisitions [10]. Low kV techniques favors safety issues on both sides – contrast medium and radiation exposure and have been widely used in pediatric CT to reduce radiation exposure without impairing image quality. In the study by Lee et al., the average DLP in neonates with CHD was 47 mGy/cm at 80 kV and 162 mGy/cm at 120 kV [11]. In our study the average DLP of all patients was 20 mGy/cm at 80 kV. Due to excessively high heart rate on infant patients, the retrospective ECG-gated scanning usually performed to acquire higher chance for good image quality. Tube current is set to prescribed value (100%) at specific phase of R–R interval and reduced to 4–20%
e794
Y. Gao et al. / European Journal of Radiology 81 (2012) e789–e795
Fig. 4. A 1-month-old boy with the diagnosis of total anomaly of pulmonary vein connection (TAPVC). Prospective ECG-triggering protocol with acquisition window of 200 ms was performed. Effective dose was 0.58 mSv. (a) Axial image showed atrial septal defect (ASD) and the right ventricle enlarged. (b) Coronal MIP and (c) volume rendered image showed the left and right pulmonary vein merged into a trunk and traveling down to the portal vein (arrow). LV, left ventricle; RV, right ventricle; PVT, pulmonary vein trunk.
at other phases. However, retrospective ECG-gated scan was associated with high radiation dose even with low tube voltage. In our study with retrospective ECG-gated spiral scan, 80 kV was used in infants less than 1 year old, tube current was modulated at full dose (100 mA) during 35–75% of the R–R interval and at 4% of the full dose for the rest of R–R interval. The average effective dose was 1.1 mSv which was significantly lower than previous study [6,12]. Recently prospectively ECG-triggering sequential data acquisition (so called “Step- and -Shoot” mode) was introduced for cardiac CTA in adults and has been employed for imaging of pediatric patients with CHD in order to reduce radiation dose [13]. Goo et al. found a low dose ECG-synchronization technique that can reduce the CT dose to 0.2–0.7 mSv in newborns and infants [12]. Cheng et al. reported the use of prospective ECG-triggering sequential CTA in pediatric patients with a mean ED of 0.38 mSv in a cohort of 35 children up to the age of 6 years [14]. To account for the higher heart rates generally observed in children the data acquisition window was centered at 40% in the end-systole [3]. In our study, prospective ECG-triggering was used with the data acquisition window of 380 ms and 200 ms centered at 45% in the cardiac cycle, tube voltage was set at 80 kV. The average ED was 0.72 mSv and 0.48 mSv for scans with 380 ms and 200 ms acquisition window. The shorter
the data acquisition window is, the lower the radiation exposure is while image quality was not affected. For diagnostic cardiac catheterizations, a median effective dose of 4.6 mSv was found for pediatric patients [15]. Therapeutic procedures resulted in a higher median effective dose of 6.0 mSv because of the prolonged use of fluoroscopy and the larger number of cine runs [15]. A study of 4952 consecutive pediatric catheterization procedures found that one or more complications occurred in 436 patients (8.8%) [1]. From our perspective, as mean radiation dose declines significantly, low-dose CTA seems better than catheter angiography in evaluating anatomical structure of children with complex CHD. 4.2. Image quality Our results indicated that using prospective ECG-triggering protocol, successfully achieved diagnostic image quality in 100% of examinations and significantly lowered radiation doses ass compared to a retrospective ECG-gated protocol. In this study, we used short-acting sedative drugs in order to minimize motion artifacts. Sleepy babies ensure good image quality with a short total examination time. We found that DSCT acquired images of diagnostic quality in all of the complex CHD patients with a very fast heart rate and with good interobserver agreement. Only minor breathing
Y. Gao et al. / European Journal of Radiology 81 (2012) e789–e795
related artifacts were observed but did not impair diagnostic image quality. For cardiovascular CT examination in infants, on the basis of 80 kV and 100 mA, using prospective ECG-triggering sequential data acquisition could reduce the radiation dose without compromising image quality. Lee et al. studied 14 neonates with CHD and found that MDCT had a diagnostic accuracy of 98%. They suggested that MDCT could replace cardiac catheterization for this indication [11]. Noninvasive CTA had shorter examination time and higher spatial resolution than magnetic resonance imaging (MRI). In comparison with cardiac catheterization, CT angiography was a safe, simple method, requiring less contrast agent and fewer complications on pediatric cardiac imaging. As the technology advances and improvements are made in scan protocols, ionizing radiation received during a CT exam will be further reduced. In our hospital almost all of the pediatric patients with complex CHD are performed with prospective ECG-triggering protocol currently. 5. Limitations Our study is a prospective randomized design for image quality and safety issues. Radiation dose has a close relation to body weight. The impact of body weight was not take into account for randomization in this study. In addition, part of infants have done conventional cardiovascular angiography. We cannot draw conclusions regarding the optimal imaging strategy in patients with complex CHD on the basis of our data. In our hospital, DSCT is used as an important second-line modality to complete or confirm diagnosis. 6. Conclusion The finding of our study shows that high spatial resolution and fast data acquisition make DSCT an obvious modality to evaluate the unstable neonate with a small heart and complex anatomy. Prospective ECG-triggering DSCT reduce radiation dose to a significantly low level while maintaining image quality. Prospective ECG-triggering DSCT angiography could be used as a very important diagnostic tool in infants with complex CHD.
e795
Conflict of interest None of the authors have any conflicting interests. References [1] Vitiello R, McCrindle BW, Nykanen D, et al. Complications associated with pediatric cardiac catheterization. Journal of the American College of Cardiology 1998;32:1433–40. [2] Jin KN, Park EA, Shin CI, et al. Retrospective versus prospective ECG-gated dual-source CT in pediatric patients with congenital heart disease: comparison of image quality and radiation dose. International Journal of Cardiovascular Imaging 2010;26:63–73. [3] Herzog C, Mulvihill DM, Nguyen SA, et al. Pediatric cardiovascular CT angiography: radiation dose reduction using automatic anatomic tube current modulation. American Journal of Radiology 2008;190:1232–40. [4] Mayo JR, Leipsic JA. Radiation dose in cardiac CT. American Journal of Radiology 2009;192:646–53. [5] Goo HW, Park IS, Ko JK, et al. CT of congenital heart disease: normal anatomy and typical pathologic conditions. RadioGraphics 2003;23:S147–65. [6] Saad MB, Rohnean A, Sigal-Cinqualbre A, et al. Evaluation of image quality and radiation dose of thoracic and coronary dual-source CT in 110 infants with congenital heart disease. Pediatric Radiology 2009;39:668–76. [7] Bamberg F, Abbaraa S, Schlett CL, et al. Predictors of image quality of coronary computed tomography in the acute care setting of patients with chest pain. European Journal of Radiology 2010;74(1):182–8. [8] Shrimpton PC. Assessment of patient dose in CT. NRPB-PE/1/2004. NRPB, Chilton. Also published as Appendix C of the 2004 CT Quality Criteria (MSCT, 2004) at: http://www.msct.eu/CT Quality Criteria.htm. [9] McCollough CH, Primak AN, Braunc N, et al. Strategies for reducing radiation dose in CT. Radiologic Clinics of North America 2009;47:27–40. [10] Sigal-Cinqualbre AB, Hennequin R, Abada HT, et al. Low-kilovoltage multidetector row chest CT in adults: feasibility and effect on image quality and iodine dose. Radiology 2004;231:169–74. [11] Lee T, Tsai IC, Fu YC, et al. Using multidetector-row CT in neonates with complex congenital heart disease to replace diagnostic cardiac catheterization for anatomical investigation: initial experiences in technical and clinical feasibility. Pediatric Radiology 2006;36:1273–82. [12] Goo HW, Seo DM, Yun TJ, et al. Coronary artery anomalies and clinically important anatomy in patients with congenital heart disease: multislice CT findings. Pediatric Radiology 2009;39:265–73. [13] Stolzmann P, Leschka S, Scheffel H, et al. Dual-source CT in step-and-shoot mode: noninvasive coronary angiography with low radiation dose. Radiology 2008;49:71–80. [14] Cheng ZP, Wang XM, Duan YH, et al. Low-dose prospective ECG-triggering dualsource CT angiography in infants and children with complex congenital heart disease: first experience. European Radiology 2010;20:2503–11. [15] Bacher K, Bogaert E, Lapere R, et al. Patient-specific dose and radiation risk estimation in pediatric cardiac catheterization. Circulation 2005;111:83–9.