International Journal of Cardiology

International Journal of Cardiology 167 (2013) 2932 2937 Contents lists available at ScienceDirect International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard Coronary artery calcium scoring: Influence of adaptive statistical iterative reconstruction using 64-MDCT Cathérine Gebhard a,1, Michael Fiechter a,b,1, Tobias A. Fuchs a, Jelena R. Ghadri a, Bernhard A. Herzog a, Felix Kuhn a, Julia Stehli a, Ennio Müller a, Egle Kazakauskaite a, Oliver Gaemperli a, Philipp A. Kaufmann a,b, a Department of Radiology, Cardiac Imaging, University Hospital Zurich, Zurich, Switzerland b Zurich Center for Integrative Human Physiology (ZIHP), University of Zurich, Zurich, Switzerland article info abstract Article history: Received 9 March 2012 Received in revised form 23 July 2012 Accepted 14 August 2012 Available online 6 September 2012 Keywords: Adaptive iterative reconstruction Coronary calcium Agatston score Cardiac computed tomography Objective: Assessment of coronary artery calcification is increasingly used for cardiovascular risk stratification. We evaluated the reliability of calcium-scoring results using a novel iterative reconstruction algorithm (ASIR) on a high-definition 64-slice CT scanner, as such data is lacking. Methods and results: In 50 consecutive patients Agatston scores, calcium mass and volume score were assessed. Comparisons were performed between groups using filtered back projection (FBP) and 20 100% ASIR algorithms. Calcium score was measured in the coronary arteries, signal and noise were measured in the aortic root and left ventricle. In comparison with FBP, use of 20%, 40%, 60%, 80%, and 100% ASIR resulted in reduced image noise between groups (7.7%, 18.8%, 27.9%, 39.86%, and 48.56%, respectively; pb0.001) without difference in signal (p=0.60). With ASIR algorithms Agatston coronary calcium scoring significantly decreased compared with FBP algorithms (837.3±130.3; 802.2±124.9, 771.5±120.7; 744.7±116.8, 724.5± 114.2, and 709.2±112.3 for 0%, 20%, 40%, 60%, 80%, and 100% ASIR, respectively, pb0.001). Volumetric score decreased in a similar manner (pb0.001) while calcium mass remained unchanged. Mean effective radiation dose was 0.81±0.08 msv. Conclusion: ASIR results in image noise reduction. However, ASIR image reconstruction techniques for HDCT scans decrease Agatston coronary calcium scores. Thus, one needs to be aware of significant changes of the scoring results caused by different reconstruction methods. 2012 Elsevier Ireland Ltd. All rights reserved. 1. Introduction Coronary artery calcifications are known to be a constituent of atherosclerosis [1 3]. As several studies proved the extent of coronary calcification to be correlated with the risk of severe cardiac events, detection and quantification of coronary calcifications by cardiac computed tomography (CT) became a useful diagnostic tool [4 6] and a risk stratification scheme for patients undergoing coronary artery calcium (CAC) screening based on the Agatston score was introduced [5]. The image reconstruction process is a fundamental determinant of image quality. Recently, Adaptive Statistical Iterative Reconstruction (ASIR, GE Healthcare) an alternative to filtered back projection (FBP) that uses iterative comparisons of each acquired projection to a synthesized projection incorporating modelling of both system optics and system statistics to reduce image noise has Financial contributions: The study was supported by grants from the Swiss National Science Foundation (SNSF) to PAK and to MF. Furthermore, we thank Patrick von Schulthess for his excellent technical support. Corresponding author at: Cardiac Imaging, University Hospital Zurich, Ramistrasse 100, NUK C 42, CH-8091 Zurich, Switzerland. Tel.: +41 44 255 41 96; fax: +41 44 255 44 14. E-mail address: pak@usz.ch (P.A. Kaufmann). 1 The first two authors contributed equally to this work. been introduced. This may permit preserved image quality with reduced tube current and lower radiation dose. A high level of reproducibility for the measurement of coronary calcification is crucial. However, in a recent study performed using a 16-MDCT scanner, it was shown that the Agatston and volumetric scores of a cardiac phantom both proved to be highly dependent on the reconstruction parameters [6]. Despite improved overall image quality, effects of different reconstruction protocols on calcium scoring [7 9], have been reported. For coronary calcium scoring there is only limited experience in choosing the optimal point of image reconstruction. Thus, we analysed the influence of ASIR versus traditional FBP on the volume score, calcium mass and Agatston score in coronary calcium measurements using an ASIR-equipped high-definition 64-slice CT (HDCT) scanner. 2. Materials and methods The study population consisted of fifty consecutive patients undergoing coronary artery calcium measurement on a 64-HDCT system (Discovery CT750 HD, GE Healthcare) within 90 days, including 19 women, with a mean age of 65.6± 10.4 years (Table 1) who were recruited after ASIR had been made available at our institution. The unenhanced CT scan for CAC scoring was obtained from the attenuation correction scan at the occasion of a myocardial perfusion scintigraphy. Thus, CAC scans were obtained from clinically indicated routine examinations. Therefore, the need for 0167-5273/$ see front matter 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijcard.2012.08.003

C. Gebhard et al. / International Journal of Cardiology 167 (2013) 2932 2937 2933 Table 1 Patient baseline and CT acquisition characteristics. Patient and CT acquisition characteristics Male sex, no of patients (of total) 31(50) Age (years), mean±sd 65.6±10.4 Weight (kg), mean±sd 77±20.4 Height (cm), mean±sd 167.9±9.3 BMI (kg/m 2 ), mean±sd 27.15±5.9 CTDI VOL (mgy) 4.24 DLP (mgy cm), mean±sd 57.6±0.9 msv, mean±sd 0.81±0.08 informed consent has been waived by the local ethics committee for the present study. Patients were retrospectively included in the study if they had signed informed consent authorizing their records to be included in our cardiac imaging research registry. Exclusion criteria were arrhythmia, prior coronary artery bypass surgery or the presence of mechanical prosthetic valves, intracoronary artery stents, pacemakers, and implantable cardioverter defibrillators. This manuscript also complies with the Principles of Ethical Publishing in the International Journal of Cardiology. 2.1. CT acquisition Coronary artery calcium scanning was performed by non-contrast cardiac CT using a 64-HDCT system. All scans were performed in cranio-caudal direction during inspiratory breathold with prospective electrocardiogram (ECG)-triggering as previously reported [10 12]. The scanning parameters included 64 0.625 mm collimation, a rotation time of 0.35 s, tube voltage of 120 kv and tube current of 200 ma. All studies were reconstructed in FBP (0%) and in 20%, 40%, 60%, 80% and 100% ASIR. 2.2. CT image reconstruction and analysis Studies were read using workstations (Advantage AW 4.4, GE Healthcare). All studies were reconstructed using ASIR ranging from 0% to 100% in 20% increments: 0% (FBP only); 20%, 40%, 60%, and 80% (composite ASIR and FBP); and 100% (ASIR only). Reconstructions were performed during postprocessing from the original data set using the scanner terminal. The signal and noise were measured in the aortic root and the left ventricle using a 0.5-cm 2 area to measure the mean signal value in Hounsfield units (HU) and SD (noise). For each image series the coronary calcium score was determined using a semiautomatic software ( SmartScore, GE Healthcare, Milwaukee, USA). All pixels with a density above a defined threshold (130 HU) were colour marked automatically. Lesions were selected manually. The software recognized the lesion size in subsequent images. Coronary artery calcium scores were separately obtained for each of the main epicardial coronary arteries left anterior descending artery (LAD), left circumflex artery (LCX), and right coronary artery (RCA) andsummed to obtain total CAC. TheLMA was assigned to the LAD vessel. From the selected areas, the software calculated the lesion volume (mm 3 ), mass (mg/cm 2 ) and Agatston score, as previously described [13 16]. 2.3. Statistical analysis Quantitative variables were expressed as mean±standard error (SEM), and categorical variables as frequencies or percentages. Comparisons between groups were performed using ANOVA for continuous variables with normal distributions and the Mann Whitney U test for continuous variables with non-normal distributions. All analyses were performed with statistics software (SPSS version 20.0 for Microsoft Windows). A two-tailed p value of b0.05 was deemed significant. For every patient the mean Agatston score, the mean volumetric calcium score, the mean calcium mass score, the signal and noise and standard deviation and error were calculated for all image reconstruction intervals separately. Normally distributed data were identified using the Shapiro Wilk test. Continuous variables are presented as mean±standard deviation and compared using an independent t-test for normally distributed data or a Mann Whitney U test for non-normally distributed data. A B C D Fig. 1. Upper panel: different percentages of iterative reconstructions at the same level of the ascending aorta (A) and the left ventricular cavity (B). Image noise expressed as the standard deviation of the attenuation (HU) in the region of interest was significantly lower in images reconstructed using iterative reconstruction compared with those reconstructed using FBP. Lower panel: different percentages of iterative reconstructions at the same level of the ascending aorta (C) and the left ventricular cavity (D). Signal to noise ratio was significantly higher in images reconstructed using iterative reconstruction compared with those reconstructed using FBP. pb0.05; pb0.001.

2934 C. Gebhard et al. / International Journal of Cardiology 167 (2013) 2932 2937 A 3. Results 3.1. Study population The study population consisted of 50 patients, including 19 women, with a mean age of 65.6±10.4 years, and a mean BMI of 27.2±5.9 kg/m 2 None of the patients received intravenous betareceptor antagonists prior to the scan. The patient baseline characteristics are listed in Table 1. 3.2. Image noise B An increased percentage of ASIR was associated with a linear reduction in noise, no change in signal, and a linear improvement in signal to noise ratio (SNR) (Fig. 1). The mean noise was 27.6±1.2, 25.5±1.2, 22.4±1.1, 19.9±1.0, 16.6±0.8 and 14.2±0.7 in the left ventricular cavity and 26.2±1.3, 23.1±1.1, 20.2±1.0, 19.7±2.2, 16.8±2.2 and 12.6±0.7 in the descending aorta for reconstructions with 0%, 20%, 40%, 60%, 80%, and 100% ASIR, respectively (pb0.001; Fig. 1A and B). Thus, as the percentage of ASIR increases the noise significantly decreases, with 100% ASIR giving the greatest noise reduction (pb0.001). There was no significant difference in the mean signal between groups (p=ns, data not shown). There was a significant increase in the SNR with increased use of ASIR in both, descending aorta and left ventricular cavity (pb0.001, Fig. 1C and D). 3.3. Calcium scoring C Fig. 2. A. Graph shows mean Agatston scores of all patients at reconstruction increments of 0, 20, 40, 60, 80 and 100% ASIR. B. Graph shows mean volumetric scores of all patients at reconstruction increments of 0, 20, 40, 60, 80 and 100% ASIR. C. Graph shows mass scores of all patients at reconstruction increments of 0, 20, 40, 60, 80 and 100% ASIR. Data are presented as mean±sd. pb0.01. 2.4. Radiation dose estimation Values for effective radiation dose were calculated by multiplying the dose length product (DLP) with a conversion factor (0.014 msv/mgy cm) as previously described [17] and adopted in large trials [18]. The use of ASIR was associated with a significant decrease in coronary calcium scores (Fig. 2). Data acquired included the Agatston score (Fig. 2A), volume score (Fig. 2B), and calcium mass (Fig. 2C). The mean of the calcium score determined by the Agatston method ranged from 837.3 to 709.2 with 100% ASIR giving the greatest reduction (pb0.001, Fig. 2A, Table 2). Using the volumetric scoring method the corresponding ranges were 318.8 to 276.9 (pb0.001, Fig. 2B, Table 2) while the range of the mass scoring method remained unchanged between 130.4 and 126.6 (p=ns, Fig. 2C, Table 2). The mean Agatston score was 837.3±130.3, 802.2±124.9, 771.5± 120.7, 744.7±116.8, 724.5±114.2, and 709.2±112.3 for reconstructions with 0%, 20%, 40%, 60%, 80%, and 100% ASIR, respectively (pb0.001; Table 2). Thus, as the percentage of ASIR increases, the Agatston significantly decreases, with 100% ASIR giving the greatest reduction (pb0.001). Similarly, the mean volumetric score was 318.8±49.1 mm 3, 304.8±46.9 mm 3, 293.4±45.1 mm 3, 289.1± 44.9 mm 3, 276±42.9 mm 3, and 276.9±43.9 mm 3 for reconstructions with 0%, 20%, 40%, 60%, 80%, and 100% ASIR, respectively (pb0.001; Table 2). In comparison with FBP, there was a decrease in Agatston score of 6.0±1.2%, 11.1±1.6%, 16.3±2.5%, 19.6±2.5%, and 22.4±2.7% for reconstructions using 20%, 40%, 60%, 80%, and 100% ASIR, respectively. Accordingly, in comparison with FBP, mean volume score was reduced by 3.7±2.4%, 8.7±2.3%, 13.2±2.7, 16.3± 2.9, and 18.6±3.0% for reconstructions using 20%, 40%, 60%, 80%, and 100% ASIR, respectively while mass scoring remained unchanged. Use of the age/sex adjusted coronary calcium score percentile allows the ranking of an individual's test result against a matched population [19,20]. However, the variable Agatston scores resulted in a different distribution of age adjusted percentiles and subsequently in a different estimation of the level of coronary risk according to percentile strata (Fig. 4B and C, comparison of FBP only versus 100% ASIR). In detail, an increased percentage of ASIR was associated with a linear reduction in calculated relative risk of 37.3±6.7%, 35.3±6.5%, 34.9±6.5%, 30.7± 6.7, and 30.3±6 for reconstructions using 20%, 40%, 60%, 80%, and 100% ASIR (pb0.001; data not shown). Moreover, variation of the coronary calcium score was significantly more pronounced in patients with low values of coronary calcifications (Agatston scoreb500) (n=30; data not shown). Maximal decrease of coronary calcium score in those

C. Gebhard et al. / International Journal of Cardiology 167 (2013) 2932 2937 2935 Table 2 Table 2 displays the mean values±sd for the volume score, calcium mass and Agatston score and their variation by percentage ASIR. Variation ofagatston score, volumetric and mass score by percentage ASIR (patients study) ASIR 0% 20% 40% 60% 80% 100% Agatston score 837.3±130.3 802.2±124.9 771.5±120.7 744.7±116.8 724.5±114.2 709.2±112.3 Volumetric score (mm 3 ) 318.8±49.1 304.8±46.9 293.4±45.1 289.1±44.9 276±42.9 216.9±43.9 Mass score (mglcm 3 ) 130.4±23.9 128.1±23.5 131.3±25 125.8±23.2 124.6±23 126.6±22.9 patients was 37.6±4.4% compared with 15.4±2.7% in patients with Agatston score >500 (Fig. 4A, n=50; pb0.001). 3.4. Radiation dose The average DLP was 57.6±0.9 mgy cm resulting in an effective radiation dose of 0.81±0.08 msv (range 0.59 0.95 msv). 4. Discussion CAC scoring has emerged as a tool for risk stratification to predict the risk for hard coronary events [21 26]. Therefore, repeat measurements should provide robust results and a low inter-scanner variability for allowing meaningful comparison during clinical follow up [14]. The rapid technical advances in multislice computed tomography and the pursuit to reduce current and radiation dose have led to the development of new reconstruction algorithms. ASIR, unlike FBP, reconstructs CT data sets by fully modelling the system statistics. ASIR makes fewer assumptions regarding the distribution of noise and utilizes an iterative process of mathematic modelling to identify and selectively reduce noise [27 29]. The significant computational power required to perform image reconstruction with ASIR has become available only recently and recent studies have shown a significant noise reduction while maintaining spatial resolution and other image quality parameters thereby permitting further reductions in current and radiation dose [27,30]. This may also play a role in CAC score scanning in view of its growing importance and based on its wide availability and applicability where contrast enhanced CT is usually avoided, for example in renal failure, arrhythmia [31,32], or known coronary total occlusion [33]. The knowledge of variability of CAC obtained with different reconstruction algorithms is crucial, but so far lacking. Probably due to the fact that these new reconstruction algorithms have been primarily developed for contrast enhanced CT imaging rather than for calcium scoring. Our study is the first to report a validation of the new ASIR algorithm in coronary calcium scoring. We confirm that the use of ASIR resulted in significant noise reduction, no change in signal, and improved SNR. However, we found consistently lower CAC scores for increasing ASIR as we found Fig. 3. Upper row shows transverse CT images reconstruction at incremental ASIR% (panel A: 0%, panel B 20%, panel C 40%). Lower row shows transverse CT images reconstruction at incremental ASIR% (panel D 60%, panel E 80%, panel F 100%). Several calcified plaques located in the proximal left anterior descending artery are clearly detectable using 0% 100% ASIR, resulting in total Agatston scores of 782, 767, 742, 712, 691, and 671, respectively. Mild smearing artefacts are seen in 40 60% ASIR whereas 100% ASIR results in a distinct underestimation of Agatston score due to severe smearing artefacts.

2936 C. Gebhard et al. / International Journal of Cardiology 167 (2013) 2932 2937 B A C borders are smoothed while the central area of the lesion appears more dense. As a result the overall calcium mass score of a lesion remains unchanged. Our study has several limitations, which have to be considered. All images were post-processed on the same workstation/software from one vendor to minimize the potential bias of different workstations/ softwares on image quality, as it has been recently shown that differences in reconstruction algorithm may introduce more variability than different scanners [38]. Therefore, the applicability of these findings to scans obtained in other types of scanner from different vendors may be limited. Finally, the median Agatston score in our study was 345, indicating that the patients represent a moderate-risk population which potentially limits the generalizability of our results. On the other hand, however, inclusion of many patients without coronary calcifications would have added only limited information to CAC comparability. We, therefore, felt it preferable including a wide range of CAC (0-3992 Agatston score). In conclusion, ASIR reconstruction algorithm above 20% is not recommendable for coronary calcium scoring. Again, this study proves that the use of ASIR is associated with an underestimation of the calcium volume and Agatston score with highest variability observed for the Agatston score. In order to allow meaningful comparison, one needs to be aware of significant changes of the scoring results caused by different reconstruction methods. Moreover, clinical outcome data and clinical recommendations are based on Agatston scores and may need to be adjusted when using ASIR. In summary, our study emphasizes the need for a standardized scan and image reconstruction protocol. Acknowledgement Fig. 4. Variation of the coronary calcium score in patients with low quantities of coronary calcifications (Agatston scoreb500) (n=30). pb0.01. B. Percentage distribution according to (age adjusted) percentiles of coronary calcium score, FBP only (=0% ASIR). C. Percentage distribution according to (age adjusted) percentiles of coronary calcium score, 100% ASIR. lower figures for the volumetric calcium score, while calcium mass scoring remained unchanged. ASIR had a statistically significant and incremental effect on CAC scoring and was associated with reduction in Agatston score of 22% in our study population. A change of at least 15 24% for Agatston score within 1 year has been suggested as a clinically meaningful progression [34,35]. Thus, in patients with low CAC scores even minimal differences in absolute values may have a great impact on clinical risk stratification and consequently on decision-making regarding downstream testing and choice of risk modifying strategies including lipid lowering drugs or platelet inhibitors [36,37]. Thus, the variability observed in the present study may represent a substantial confounding bias. Moreover, absolute precision of repeat measurements appears most relevant in patients without excessive calcifications, in whom risk factor modification may slow progression and therefore its monitoring may be appropriate. The use of 20% and 40% ASIR demonstrated the least variation of the Agatston score, but a significant noise reduction. Use of 20%, 40% and 60% ASIR led to a moderate number of patients who had to be reassigned to other cardiovascular risk groups while with 80% and 100% ASIR a substantial fraction of 18% of patients had to be re-assigned to a lower risk group (Fig. 3). While ASIR had a significant influence on the volumetric score and the Agatston score, mass scoring remained unaffected and did not change with increasing percentage of ASIR used for reconstruction. This might be due to a lower susceptibility of the mass score to partial volume effects. Reconstructions with high proportions of ASIR yield images which differ from those obtained with traditional FBP as the The authors of this manuscript have certified that they comply with the Principles of Ethical Publishing in the International Journal of Cardiology. References [1] Arad Y, Goodman KJ, Roth M, Newstein D, Guerci AD. Coronary calcification, coronary disease risk factors, C-reactive protein, and atherosclerotic cardiovascular disease events: the St. Francis Heart Study. J Am Coll Cardiol 2005;46:158-65. [2] Esteves FP, Khan A, Correia LC, et al. Absent coronary artery calcium excludes inducible myocardial ischemia on computed tomography/positron emission tomography. Int J Cardiol 2011;147:424-7. [3] Lim S, Choi HJ, Shin H, et al. Subclinical atherosclerosis in a community-based elderly cohort: The Korean Longitudinal Study on Health and Aging. Int J Cardiol 2012;155:126-33. [4] Versteylen MO, Joosen IA, Winkens MH, et al. Combined use of exercise electrocardiography, coronary calcium score and cardiac CT angiography for the prediction of major cardiovascular events in patients presenting with stable chest pain. Int J Cardiol 2013 July 15;167(1):121-5. [5] Agatston AS, Janowitz WR, Hildner FJ, Zusmer NR, Viamonte Jr M, Detrano R. Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol 1990;15:827-32. [6] Begemann PG, van Stevendaal U, Koester R, et al. Evaluation of the influence of acquisition and reconstruction parameters for 16-row multidetector CT on coronary calcium scoring using a stationary and dynamic cardiac phantom. Eur Radiol 2007;17:1985-94. [7] Schlosser T, Hunold P, Schmermund A, et al. Coronary artery calcium score: influence of reconstruction interval at 16-detector row CT with retrospective electrocardiographic gating. Radiology 2004;233:586-9. [8] Schlosser T, Hunold P, Voigtlander T, Schmermund A, Barkhausen J. Coronary artery calcium scoring: influence of reconstruction interval and reconstruction increment using 64-MDCT. AJR Am J Roentgenol 2007;188:1063-8. [9] Mahnken AH, Wildberger JE, Sinha AM, et al. Variation of the coronary calcium score depending on image reconstruction interval and scoring algorithm. Invest Radiol 2002;37:496-502. [10] Ghadri JR, Kuest SM, Goetti R, et al. Image quality and radiation dose comparison of prospectively triggered low-dose CCTA: 128-slice dual-source high-pitch spiral versus 64-slice single-source sequential acquisition. Int J Cardiovasc Imaging 2012;28(5):1217-25. [11] Pontone G, Andreini D, Bartorelli AL, et al. Radiation dose and diagnostic accuracy of multidetector computed tomography for the detection of significant coronary artery stenoses: A meta-analysis. Int J Cardiol 2012 Oct 18;160(3):155-64.

C. Gebhard et al. / International Journal of Cardiology 167 (2013) 2932 2937 2937 [12] Pontone G, Andreini D, Bartorelli AL, et al. Diagnostic accuracy of coronary computed tomography angiography: a comparison between prospective and retrospective electrocardiogram triggering. J Am Coll Cardiol 2009;54:346-55. [13] McCollough CH, Ulzheimer S, Halliburton SS, Shanneik K, White RD, Kalender WA. Coronary artery calcium: a multi-institutional, multimanufacturer international standard for quantification at cardiac CT. Radiology 2007;243:527-38. [14] Oudkerk M, Stillman AE, Halliburton SS, et al. Coronary artery calcium screening: current status and recommendations from the European Society of Cardiac Radiology and North American Society for Cardiovascular Imaging. Int J Cardiovasc Imaging 2008;24:645-71. [15] Ueda H, Harimoto K, Tomoyama S, et al. Association between cardiovascular risk factors and the presence of coronary plaque in a zero or low coronary artery calcium score. Int J Cardiol 2011;147:475-7. [16] Morita H, Fujimoto S, Kondo T, et al. Prevalence of computed tomographic angiography-verified high-risk plaques and significant luminal stenosis in patients with zero coronary calcium score. Int J Cardiol 2012;158:272-8. [17] Hausleiter J, Meyer T, Hermann F, et al. Estimated radiation dose associated with cardiac CT angiography. JAMA 2009;301:500-7. [18] Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307-10. [19] Hoff JA, Chomka EV, Krainik AJ, Daviglus M, Rich S, Kondos GT. Age and gender distributions of coronary artery calcium detected by electron beam tomography in 35,246 adults. Am J Cardiol 2001;87:1335-9. [20] Kondos GT, Hoff JA, Sevrukov A, et al. Electron-beam tomography coronary artery calcium and cardiac events: a 37-month follow-up of 5635 initially asymptomatic low- to intermediate-risk adults. Circulation 2003;107:2571-6. [21] Ghadri JR, Pazhenkottil AP, Nkoulou RN, et al. Very high coronary calcium score unmasks obstructive coronary artery disease in patients with normal SPECT MPI. Heart 2011;97:998 1003. [22] Detrano R, Guerci AD, Carr JJ, et al. Coronary calcium as a predictor of coronary events in four racial or ethnic groups. N Engl J Med 2008;358:1336-45. [23] Budoff MJ, Shaw LJ, Liu ST, et al. Long-term prognosis associated with coronary calcification: observations from a registry of 25,253 patients. J Am Coll Cardiol 2007;49:1860-70. [24] Groothuis JG, Beek AM, Meijerink MR, et al. Positive predictive value of computed tomography coronary angiography in clinical practice. Int J Cardiol 2012;156(3): 315-9. [25] Dedic A, Rossi A, Ten Kate GJ, et al. First-line evaluation of coronary artery disease with coronary calcium scanning or exercise electrocardiography. Int J Cardiol 2013 Feb 20;163(2):190-5. [26] Sosnowski M, Pysz P, Szymanski L, Gola A, Tendera M. Negative calcium score and the presence of obstructive coronary lesions in patients with intermediate CAD probability. Int J Cardiol 2011;148:e16-8. [27] Leipsic J, Labounty TM, Heilbron B, et al. Estimated radiation dose reduction using adaptive statistical iterative reconstruction in coronary CT angiography: the ERASIR study. AJR Am J Roentgenol 2010;195:655-60. [28] Sagara Y, Hara AK, Pavlicek W, Silva AC, Paden RG, Wu Q. Abdominal CT: comparison of low-dose CT with adaptive statistical iterative reconstruction and routinedose CT with filtered back projection in 53 patients. AJR Am J Roentgenol 2010;195: 713-9. [29] Hara AK, Paden RG, Silva AC, Kujak JL, Lawder HJ, Pavlicek W. Iterative reconstruction technique for reducing body radiation dose at CT: feasibility study. AJR Am J Roentgenol 2009;193:764-71. [30] Pontone G, Andreini D, Bartorelli AL, et al. Feasibility and diagnostic accuracy of a low radiation exposure protocol for prospective ECG-triggering coronary MDCT angiography. Clin Radiol 2012;67:207-15. [31] Tsiflikas I, Drosch T, Brodoefel H, et al. Diagnostic accuracy and image quality of cardiac dual-source computed tomography in patients with arrhythmia. Int J Cardiol 2010;143:79-85. [32] Uehara M, Funabashi N, Takaoka H, Komuro I. Quality of coronary arterial 320-slice computed tomography images compared with 16-slice computed tomography images in subjects with chronic atrial fibrillation. Int J Cardiol 2011;149:e90-3. [33] Hsu JT, Kyo E, Chu CM, Tsuji T, Watanabe S. Impact of calcification length ratio on the intervention for chronic total occlusions. Int J Cardiol 2011;150:135-41. [34] Raggi P, Cooil B, Shaw LJ, et al. Progression of coronary calcium on serial electron beam tomographic scanning is greater in patients with future myocardial infarction. Am J Cardiol 2003;92:827-9. [35] Oudkerk M, Stillman AE, Halliburton SS, et al. Coronary artery calcium screening: current status and recommendations from the European Society of Cardiac Radiology and North American Society for Cardiovascular Imaging. Eur Radiol 2008;18:2785-807. [36] Rozanski A, Gransar H, Shaw LJ, et al. Impact of coronary artery calcium scanning on coronary risk factors and downstream testing the EISNER (Early Identification of Subclinical Atherosclerosis by Noninvasive Imaging Research) prospective randomized trial. J Am Coll Cardiol 2011;57:1622-32. [37] Uretsky S, Rozanski A, Supariwala A, et al. Clinical outcomes following a strategy of optimized medical management and selective "downstream" procedures following coronary computed tomography angiography. Int J Cardiol 2013 May 25;165(3):468-73. [38] Ghadri JR, Goetti R, Fiechter M, et al. Inter-scan variability of coronary artery calcium scoring assessed on 64-multidetector computed tomography vs. dual-source computed tomography: a head-to-head comparison. Eur Heart J 2011;32:1865-74.