- Email: [email protected]
Radiation doses from phantom measurements at high-pitch dual-source computed tomography coronary angiography
European Journal of Radiology 81 (2012) 773–779
Contents lists available at ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Radiation doses from phantom measurements at high-pitch dual-source computed tomography coronary angiography Robert Goetti a,∗ , Sebastian Leschka b , Markus Boschung c , Sabine Mayer c , Christophe Wyss d , Paul Stolzmann a,e , Thomas Frauenfelder a a
Department of Diagnostic Radiology, University Hospital Zurich, Raemistrasse 100, CH-8091 Zurich, Switzerland Institute of Radiology, Kantonsspital St. Gallen, Rorschacherstrasse 95, CH-9007 St. Gallen, Switzerland c Division for Radiation Safety and Security, Paul Scherrer Institut, CH-5232 Villigen, Switzerland d Division of Cardiology, University Hospital Zurich, Raemistrasse 100, CH-8091 Zurich, Switzerland e Cardiac MR PET CT Program, Massachusetts General Hospital and Harvard Medical School, 165 Cambridge St, Boston, MA 02114, USA b
a r t i c l e
i n f o
Article history: Received 14 September 2010 Received in revised form 6 January 2011 Accepted 17 January 2011 Keywords: Radiation dose High-pitch Computed tomography Coronary angiography Anthropomorphic phantom
a b s t r a c t Objective: To compare radiation doses delivered at prospectively ECG-triggered sequential- (SEQ), retrospectively ECG-gated spiral- (RETRO) and prospectively ECG-triggered high-pitch spiral- (HP) computed tomography coronary angiography (CTCA) protocols, as well as catheter coronary angiography (CCA) using an anthropomorphic phantom. Materials and methods: An anthropomorphic Alderson phantom equipped with 50 thermoluminescent dosimeters (TLDs) was scanned using different CTCA protocols and an uncomplicated diagnostic CCA examination was simulated. Absorbed doses were experimentally determined and effective doses calculated using the dose-length product (DLP) for CTCA and the dose-area product (DAP) for CCA, as well as according to International Commission on Radiation Protection (ICRP) publications 60 and 103. Results: Effective organ doses were significantly lower for HP protocols (100 kV: 0.17 ± 0.26 mSv; 120 kV: 0.26 ± 0.39 mSv) compared to SEQ protocols (100 kV: 0.50 ± 0.79 mSv; 120 kV: 0.90 ± 1.41 mSv; each p < 0.05) and compared to RETRO protocols (100 kV: 1.59 ± 2.12 mSv; 120 kV: 2.75 ± 3.50 mSv; each p < 0.05). Effective organ doses at HP-CTCA tended to be lower than at CCA (0.37 ± 0.40 mSv), however this was not statistically significant (p = 0.13). Effective doses calculated according to ICRP guidelines could be estimated using the DLP and a conversion coefficient of k = 0.034 mSv/[mGy cm] (ICRP103) or k = 0.028 mSv/[mGy cm] (ICRP60), respectively. HP-CTCA led to a dose reduction of 89% compared to RETRO-CTCA, regardless of the calculation method used. Conclusions: Radiation doses as determined by phantom measurements are significantly lower at HPCTCA compared to SEQ-CTCA and RETRO-CTCA and comparable to uncomplicated diagnostic CCA. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Computed tomography coronary angiography (CTCA) has emerged as a robust non-invasive alternative to catheter coronary angiography (CCA) for the diagnosis of coronary artery disease (CAD) [1,2]. However with the introduction of 16- and 64-slice CT systems, rising radiation doses – due to the necessity of higher tube current at higher spatial resolution [3] – have become a concern with reported doses of up to 30 mSv [4] when applying retrospectively ECG-gated acquisition protocols at a low pitch of 0.2–0.5. This concern has led to the development of various strategies to reduce radiation exposure to the patient, including ECG-based tube current modulation [5], tube voltage reduction in slim patients [6],
∗ Corresponding author. Tel.: +41 44 255 11 11; fax: +41 44 255 44 43. E-mail address: [email protected] (R. Goetti). 0720-048X/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2011.01.068
use of sequential scanning [5,7], volume scanning in 320-detector single-source CT [3], and the most recently introduced high-pitch spiral scanning in dual-source CT [2]. Patient radiation dose estimates at CTCA are usually derived from scanner protocol data and a method proposed by the European Working Group for Guidelines on Quality Criteria for CT [8] by multiplication of the dose-length-product (DLP) with a region-specific conversion coefficient. Furthermore, commercially available software applications allow more accurate and individualized estimates of effective doses delivered to different organs by specification of patient parameters such as gender, size, exact extent of scanned body region as well as scanning parameters such as tube voltage, tube current, slice collimation and pitch [9,10]. However, these methods do not actually measure radiation doses delivered to different organs. It has been reported by various authors [3,11] that effective doses estimated using the DLP may be substantially lower than those calculated according to the guide-
774
R. Goetti et al. / European Journal of Radiology 81 (2012) 773–779
Fig. 1. Photograph of the Alderson phantom used for the study (left) with corresponding CT scanogram (right) showing the embedded skeleton and different attenuation of lungs and soft tissue. The box within the CT scanogram image represents the 13.5 cm scan range.
lines of the International Commission on Radiation Protection (ICRP), and vary between calculations according to ICRP publications 60 [12] and 103 [13], due to different tissue weighting factors. Whereas effective doses can be estimated in patient studies using the DLP and a conversion coefficient, effective dose calculations according to ICRP publication 60 and 103 guidelines are obtained by phantom measurements. Such measurements have been reported in various studies for CTCA using retrospectively ECG-gated spiral-, prospectively ECG-triggered sequential and volume acquisition protocols [3,5,11]. To the best of our knowledge, there is however no study in which effective doses at CTCA performed in the high-pitch spiral acquisition mode were calculated from experimentally determined doses and compared to those of conventional CTCA protocols and catheter coronary angiography. The purpose of this study therefore was measure and compare effective doses delivered at CTCA in conventional (i.e. retrospectively ECG-gated spiral acquisition and prospectively ECG-triggered sequential acquisition) and prospectively ECG-triggered high-pitch spiral acquisition protocols, as well as catheter coronary angiography using an anthropomorphic Alderson phantom.
2. Materials and methods 2.1. Dosimetry An anthropomorphic male Alderson phantom was used for all dose measurements (Fig. 1). This phantom consists of a head, neck, and torso, without arms and legs, simulating a male of 175 cm height and 73.5 kg weight. It includes a human skeleton embedded in isocyanide rubber mass (specific density: 0.985 g/cm3 , mean atomic number 7.3) simulating soft tissue, and contains preformed lung substitutes consisting of microcellular foam (specific density: 0.32 g/cm3 , mean atomic number 7.3). It is formed of 35 slices of 2.5 cm thickness, each slice containing a grid of multiple boreholes (spaced by 3 cm from each other) in which thermoluminescent dosimeters (TLDs) can be placed.
Lithium fluoride TLDs (TLD-700; Thermo Fisher Scientific Inc, Waltham, Massachusetts) were used in this study. The dimensions of these dosimeters are 3.2 × 3.2 × 0.9 mm3 . All evaluations of the TLDs were performed with a commercially available reader (Harshaw 5500, Thermo Fisher Scientific Inc, Waltham, Massachusetts). For stabilization, the TLDs were put in an oven for 1 h at a temperature of 90 ◦ C prior to the evaluation. After each evaluation the TLDs were annealed at 400 ◦ C for 1 h. The TLD signal (in nC) was corrected for the individual sensitivity of the TLD and multiplied by a calibration factor to obtain absorbed dose in water (in mGy). For each examination 50 TLDs were positioned inside the phantom. Ten TLDs were used as blanks. A calibration factor was determined in parallel for each examination by irradiating 10 TLDs free-in-air in a reference field of a 137 Cs source. The air kerma at the calibration position was measured with an ionization chamber and converted into absorbed dose in water. Typical calibration factors obtained were 0.4 mGy/nC. The overall uncertainty of the measured doses is about 9%, including the uncertainty of the individual sensitivity (7%) and the uncertainty of the calibration factor (5%). The 50 TLDs used at each examination were positioned at the location of different organs as previously described [11]: bone marrow, esophagus, thyroid gland, lungs, breasts, skin, stomach, liver, colon, ovaries, bladder and testicles. The numbers and locations of the TLDs in the respective slices of the Alderson phantom and according organs are listed in Table 1. Effective doses were calculated from the TLD measurements according to the guidelines published in the International Commission on Radiological Protection (ICRP) Publications 60 and 103 [12,13] by multiplying the mean experimentally determined dose of each organ with the corresponding tissue weighting factors published in ICRP publications 60 and 103. Organ doses of tissues with individual tissue weighting factors in both ICRP 60 and 103 were determined by calculating the mean of at least two (range, 2–12, see Table 1) measurements of TLDs positioned at corresponding locations of the phantom. Doses of remainder tissues were averaged from measurements of adjacent organs. As the TLDs were calibrated in absorbed dose in water, the organ doses were corrected with the organ specific mass energy-absorption coefficient. The uncertainty
R. Goetti et al. / European Journal of Radiology 81 (2012) 773–779 Table 1 Location and number of thermoluminescent dosimeters (TLDs) within the slices of the Alderson phantom. Phantom slice number
Number of TLDs (n = 50)
Organs
3 9 13 14 17 18 21 26 28 29 31 32 34
2 4 (2 + 2)a 6 (2 + 4) 2 6 (2 + 4) 6 (2 + 4) 6 (2 + 2 + 2) 2 4 (2 + 2) 2 4 2 4
Skull (bone marrow) Esophagus, thyroid gland Ribs (bone marrow), lungs Breasts Ribs (bone marrow), lungs Spine (bone marrow), skin Esophagus, stomach, liver Colon Pelvis (bone marrow), colon Sacrum (bone marrow) Ovaries Bladder Testicles
a Numbers in brackets indicate the numbers of TLDs for each organ in the respective slice.
of the effective dose introduced by this correction is in the order of 20%. 2.2. Computed tomography coronary angiography Eight different CTCA protocols were evaluated (Table 2): (a) two retrospectively ECG-gated spiral (RETRO) acquisition protocols (at 100 kV and 120 kV tube voltage, scan range 15 cm) (b) two prospectively ECG-triggered sequential “step-and-shoot” (SEQ) acquisition protocols (at 100 kV and 120 kV tube voltage, scan range 13.5 cm) (c) four prospectively ECG-triggered high-pitch spiral (HP) acquisition protocols (at 100 kV and 120 kV tube voltage; 13.5 cm and 15 cm scan range). All scans were performed on a second-generation dual-source CT system (Somatom Definition Flash, Siemens Healthcare, Forchheim, Germany). Tube current-time product was set at 320 mAs for all scans; tube voltages were 100 kV or 120 kV; pitch was 0.23 for retrospectively ECG-gated spiral scans and 3.2 for prospectively ECG-triggered spiral scans. The scan ranges were 13.5 cm or 15 cm, approximately corresponding to the 50th and 75th percentiles of scan ranges of CTCA reported in a recent multicenter trial [4]. The 15 cm scan range for RETRO acquisitions was chosen to begin at a level 2 cm below the tracheal bifurcation and end at the apex of the heart of the anthropomorphic Alderson phantom. For SEQ acquisitions, scan ranges can only be set in fixed intervals. Thus, four blocks of sequential detector coverage were chosen, corresponding to a scan range of 13.5 cm, with the same center of the scan range as in RETRO acquisitions. To assure comparability of dose values, HP acquisitions were performed at both 13.5 cm and 15 cm
775
scan ranges and all comparisons were made only between acquisitions performed with the same scan range. A heart rate of 60 beats per minute was simulated by the scanner software for all scans and full tube current was applied from 30 to 80% of the RR-interval for retrospectively ECG-gated spiral scans and from 60 to 80% of the RR-interval for prospectively ECG-triggered sequential scans to simulate clinically established protocols [14,15]. For prospectively triggered high-pitch scans, 60% of the RR-interval was set to trigger the scan and full tube current was applied during the entire scan. To achieve a sufficient exposure of the TLDs for accurate measurement of absorbed doses, high-pitch CTCA scans were performed ten times and all other CTCA scans were performed five times. The resulting doses were divided by ten or five, respectively, to calculate the reported single-scan experimentally determined doses. Effective doses at CTCA were estimated using the method proposed by the European Working Group for Guidelines on Quality Criteria for CT [8] using the dose-length product (DLP) and a conversion coefficient for the chest (k = 0.017 mSv/[mGy cm]) and were calculated from the experimentally determined doses according to the guidelines in ICRP publications 60 and 103 as described above. To compare image quality between the different protocols, image noise was measured as follows: for all datasets (HP, SEQ and RETRO, each at 100 kV and 120 kV) images were reconstructed using standard cardiac imaging parameters (slice thickness 0.75 mm, increment 0.5 mm, convolution kernel B26f). Regions of interest (ROIs) of 6 cm2 – avoiding boreholes – were drawn at five different locations (ascending aorta, aortic root, basal heart, midventricular heart and apical heart) in each image set. Noise was defined as the standard deviation from the attenuation measurements in these ROIs. All measurements were performed by two radiologists blinded to the acquisition mode.
2.3. Catheter coronary angiography For the simulation of a diagnostic CCA examination, a biplane angiographic system was used (Integris Allura 9, Philips Medical Systems, Best, The Netherlands). Exposure times for an uncomplicated diagnostic examination were determined and the examination was simulated by an experienced interventional cardiologist (with seven years of experience in interventional cardiology, anonymized) as follows: firstly, 10 s anterior–posterior (AP) fluoroscopy to simulate the intubation of the left coronary artery (LCA) and 10 s 60◦ left anterior oblique (LAO 60) fluoroscopy simulating the intubation of the right coronary artery (RCA). Then, the simulation of LCA angiography was performed with three exposures in six orientations and RCA angiography with two exposures in four orientations. Each exposure consisted of 1 s fluoroscopy and 2 s cine-imaging. Detailed exposure parameters of the examination are provided in Table 3. In order to achieve a sufficient exposure of the TLDs for accurate measurement of effective doses, the entire examination was performed five times and the total dose divided
Table 2 Computed tomography coronary angiography protocols. Protocol namea
ECG-synchronisation
Acquisition
Tube voltage (kV)
Tube current (mAs)
ECG-pulsing (% of RR-interval)
Pitch
CTDIvol (mGy)
Scan length (mm)
HP100S HP120S HP100L HP120L SEQ100 SEQ120 RETRO100 RETRO120
Prospective Prospective Prospective Prospective Prospective Prospective Retrospective Retrospective
Spiral Spiral Spiral Spiral Sequential Sequential Spiral Spiral
100 120 100 120 100 120 100 120
320 320 320 320 320 320 320 320
– – – – 60–80% 60–80% 30–80% 30–80%
3.2 3.2 3.2 3.2 – – 0.23 0.23
3.10 5.18 3.10 5.18 14.14 23.65 34.53 60.81
135 135 150 150 135 135 150 150
a HP: high-pitch; SEQ: sequential; RETRO: retrospectively ECG-gated; 100: 100 kV tube voltage; 120: 120 kV tube voltage; S: short scan range (135 mm); L: long scan range (150 mm).
776
R. Goetti et al. / European Journal of Radiology 81 (2012) 773–779
Table 3 Orientations, tube voltages and tube currents for the three left coronary artery exposures and two right coronary artery exposures of simulated catheter coronary angiography. Exposure
Orientations
Tube voltage (kV)
Tube current (mAs)
LCA 1
AP LAO90
67 73
439 751
LCA 2
RAO30 CAUD20 LAO40 CRAN20
69
519
73
727
RAO15 CRAN30 LAO50 CAUD20
72
681
74
782
RCA 1
RAO30 LAO30
69 75
547 841
RCA 2
LAO15 CRAN30 LAO90
68
495
74
782
LCA 3
LCA 1–3: left coronary artery exposures; AP: anterior–posterior; LAO90: left anterior oblique, 90◦ angle; RAO30: right anterior oblique, 30◦ angle; CAUD20: 20◦ caudal angulation, CRAN20: 20◦ cranial angulation; RCA 1–2: right coronary artery exposures.
by five. Effective doses at CCA were estimated by multiplying the dose-area product (DAP) of the CCA examination with a conversion coefficient for the chest (k = 0.22 mSv/[Gy cm2 ]) [16] and were calculated from the TLD measurements according to the guidelines in ICRP publications 60 and 103 as described above. 2.4. Statistical analyses All statistical analyses were performed using commercially available software (PASW Statistics 18, SPSS Inc., Chicago, IL). Continuous variables are expressed as means ± standard deviations and categorical variables are expressed as frequencies or percentages. Normality testing was performed using the Kolmogorov–Smirnov test. Differences in experimentally determined organ doses between the CTCA and CCA protocols and differences in experimentally determined organ doses between CTCA protocols at 100 kV and 120 kV tube voltage as well as differences between effective doses calculated according to ICRP 60 and 103 guidelines and DLP were assessed using the paired-samples t-test. Correlations between effective doses at CTCA calculated according to ICRP 60 and 103 guidelines and DLP were assessed using Pearson’s correlation coefficient and linear regression analysis. Bland–Altman analysis was used to assess differences in effective doses between calculations according to ICRP and using the DLP (for CTCA) or using the DAP (for CCA). Differences in image noise measurements between the different CTCA protocols were evaluated with the paired samples t-test. For all tests, a two-tailed p < 0.05 was considered to indicate statistical significance. 3. Results 3.1. Computed tomography coronary angiography Table 4 provides an overview of experimentally determined doses (mGy) as well as calculated effective doses (mSv) for the different protocols. Experimentally determined as well as calculated effective doses were distributed normally according to the Kolmogorov–Smirnov test and were therefore compared using paired samples t-tests. The highest effective organ doses (mSv) as calculated according to ICRP103 guidelines from experimentally determined doses at CTCA were delivered to
Fig. 2. Scatterplots demonstrating excellent correlation between effective doses of the eight CTCA protocols calculated according to the guidelines of the International Commission on Radiation Protection (ICRP) publications 60 (triangles) and 103 (circles) and the dose-length product (DLP). Dotted lines indicate 95% confidence intervals.
the breast, followed by the lungs. Effective organ doses were significantly lower for HP protocols (100 kV: 0.17 ± 0.26 mSv; 120 kV: 0.26 ± 0.39 mSv) compared to SEQ protocols (100 kV: 0.50 ± 0.79 mSv; 120 kV: 0.90 ± 1.41 mSv; each p < 0.05) when using the same scan length of 13.5 cm, and compared to RETRO protocols (100 kV: 0.19 ± 0.26 mSv vs. 1.59 ± 2.12 mSv; 120 kV: 0.29 ± 0.40 mSv vs. 2.75 ± 3.50 mSv; each p < 0.05) when using the same scan length of 15 cm. Experimentally determined organ doses using 100 kV protocols were significantly lower than those at 120 kV protocols (each p < 0.05). Effective doses delivered at CTCA as calculated according to the ICRP guidelines showed a wide range from 1.8 mSv to 35.8 mSv depending on the applied protocol and the different tissue weighting factors for calculations according to ICRP 60 and ICRP 103. Bland–Altman analysis revealed a higher difference between effective doses as estimated using the DLP and effective dose as calculated according to ICRP 103 guidelines (5.4 ± 6.0 mSv) than between effective doses estimated using the DLP and calculated according to ICRP 60 (3.4 ± 3.9 mSv). Effective doses as calculated according to ICRP 103 guidelines were 23.5 ± 3.1% higher than those calculated according to the ICRP 60 guidelines, and 103.4 ± 8.2% higher than those estimated using the DLP and a conversion coefficient of k = 0.017 mSv/[mGy cm] (each p < 0.05). There was however an excellent correlation between effective doses calculated according to ICRP 60 and 103 guidelines and DLP (R2 = 0.999, R2 = 1; each p < 0.001) for the different CTCA protocols (Fig. 2). Linear regression analysis revealed that effective doses according to ICRP calculations could be estimated using the DLP provided in the patient protocol of this specific scanner and a conversion coefficient of k = 0.034 mSv/[mGy cm] (ICRP103) or k = 0.028 mSv/[mGy cm] (ICRP60), respectively. Using the same scan range of 15 cm and a tube current of 120 kV, applying the prospectively ECG-triggered high-pitch spiral protocol led to a dose reduction of 89% compared to the retrospectively ECG-gated protocol, regardless of the calculation method used. There was no significant difference in image noise between
Table 4 Experimentally determined and effective doses at CTCA and CCA. Protocol a
Organ
HP100S
HP100L
HP120S
HP120L
Bone marrow Esophagus Thyroid Lungs Breast Skin Stomach Liver Colon Ovaries Testicles Bladder
1.56 3.11 0.48 5.35 6.73 4.78 1.91 1.77 0.10 0.02 0.01 0.02
1.71 2.68 0.47 4.97 6.82 4.53 3.11 2.83 0.13 0.03 0.02 0.02
2.79 5.19 1.07 9.22 9.09 8.34 2.55 2.19 0.17 0.04 0.03 0.03
3.01 4.54 0.83 8.71 9.86 7.51 4.53 5.11 0.25 0.05 0.03 0.04
61 –
DLP (mGy cm) DAP (mGy cm2 ) Effective dose (mSv)
From DLPb From DAPc ICRP 60 ICRP 103
1.0 – 1.8 2.2
65 – 1.1 – 1.9 2.3
102 – 1.7 – 2.8 3.4
109 – 1.9 – 3.2 3.9
SEQ100 4.83 11.07 1.10 14.84 21.49 18.13 3.84 4.13 0.27 0.07 0.06 0.05 195 – 3.3 – 5.0 6.3
SEQ120
RETRO100
RETRO120
CCA
8.39 19.11 1.98 25.89 38.66 33.79 8.60 7.06 0.57 0.14 0.08 0.10
15.38 29.77 2.49 40.83 57.41 55.91 23.31 27.88 1.33 0.25 0.13 0.17
27.46 51.32 4.77 72.07 87.70 89.93 51.11 42.97 2.79 0.58 0.26 0.38
7.49 8.08 0.13 3.71 1.59 12.60 9.26 22.88 1.05 0.16 0.03 0.09
326 – 5.5 – 9.1 11.7
603 – 10.3 – 16.6 20.6
1043 – 17.7 – 29.5 35.8
– 14,313 – 3.1 4.6 4.8
HP100S: prospectively ECG-triggered high-pitch spiral acquisition at 100 kV with 13.5 cm scan range; HP100L: prospectively ECG-triggered high-pitch spiral acquisition at 100 kV with 15 cm scan range; HP120S: prospectively ECG-triggered high-pitch spiral acquisition at 120 kV with 13.5 cm scan range; HP120L: prospectively ECG-triggered high-pitch spiral acquisition at 120 kV with 15 cm scan range; SEQ100: prospectively ECG-triggered sequential acquisition at 100 kV with 13.5 cm scan range; SEQ120: prospectively ECG-triggered sequential acquisition at 120 kV with 13.5 cm scan range; RETRO100: retrospectively ECG-gated spiral acquisition at 100 kV with 15 cm scan range; RETRO120: retrospectively ECG-gated spiral acquisition at 120 kV with 15 cm scan range. a Only organs with individually specified tissue weighting factors both in ICRP publications 60 and 103 are shown. b Effective dose calculated for CTCA by multiplying the DLP with a conversion coefficient of k = 0.017 mSv/[mGy cm]. c Effective dose calculated for CCA by multiplying the DAP with a conversion coefficient of k = 0.22 mSv/[Gy cm2 ].
R. Goetti et al. / European Journal of Radiology 81 (2012) 773–779
Experimentally determined dose (mGy)
777
778
R. Goetti et al. / European Journal of Radiology 81 (2012) 773–779
the measurements performed by the two readers (15.1 ± 2.2 vs. 15.0 ± 2.3, p = 0.33), indicating good inter-reader agreement. Image noise was significantly lower in 120 kV protocols than in 100 kV protocols (14.0 ± 1.2 vs. 16.2 ± 2.5, p < 0.001). There were no significant differences in image noise between the SEQ and RETRO protocols at 100 kV (14.7 ± 1.6 vs. 14.8 ± 1.6, p = 0.87) and 120 kV (13.4 ± 0.6 vs. 13.7 ± 1.1, p = 0.22), respectively. However, HP protocols (100 kV: 19.0 ± 1.5; 120 kV: 14.8 ± 1.2) yielded significantly higher image noise compared to SEQ and RETRO acquisitions at the same tube voltages (all p < 0.05). Image noise in HP acquisition at 120 kV was comparable to image noise in SEQ acquisition at 100 kV (p = 0.72) whereas experimentally determined doses were significantly lower using the HP protocol at 120 kV compared to the SEQ protocol at 100 kV (3.4 ± 3.6 mGy vs. 6.7 ± 7.7 mGy, p < 0.05). 3.2. Catheter coronary angiography At CCA, the highest effective doses as calculated from experimentally determined doses according to ICRP guidelines were delivered at the upper abdominal organs stomach and liver (Table 4). The effective dose calculated for CCA was 4.6 mSv (ICRP 60) or 4.8 mSv (IRCP 103), which are 48.4–54.8% higher than the effective dose estimated by calculation using the DAP. 3.3. CTCA vs. CCA Effective doses calculated according to ICRP guidelines were lower in the four different HP-CTCA protocols (range, 1.8–3.9 mSv) compared to those at CCA (range, 4.6–4.8 mSv), however the differences in effective organ doses did not reach statistical significance (CCA: 0.37 ± 0.40 mSv; HP-CTCA, 100 kV, 13.5 cm scan range: 0.17 ± 0.26 mSv, p = 0.13; HP-CTCA, 120 kV, 15 cm scan range: 0.29 ± 0.40 mSv, p = 0.57). The highest effective organ doses were delivered to the breast and lungs at CTCA and to the stomach and liver at CCA. 4. Discussion The main findings of our study are: 1. High-pitch CTCA protocols are associated with significantly lower effective doses, than both prospectively ECG-triggered sequential protocols and retrospectively ECG-gated spiral protocols with potential dose reductions of up to 89% and effective doses comparable to those delivered at uncomplicated diagnostic CCA. 2. Effective dose estimates vary substantially depending on the method of calculation for both CTCA and CCA and are significantly different between calculations according to ICRP guidelines and calculations using the DLP and a conversion coefficient of k = 0.017 mSv/[mGy cm] for CTCA, however with an excellent correlation between DLP and ICRP calculations. The wide range of radiation doses at CTCA reported for different protocols and at different centers is an issue of concern pointed out by various authors [4,17] with a reported range of median effective doses varying between 5 mSv and 30 mSv [4]. Our results for SEQ and RETRO acquisition protocols are comparable to those reported in a recent dual-source CTCA phantom study, with effective doses ranging from 2.8 mSv to 32.6 mSv – depending on gender, heart rate acquisition protocol and ECG-pulsing – when taking into account that those measurements were made using a shorter scan range, variable low pitch and different ECG-pulsing windows in a firstgeneration dual-source scanner [5]. Applying the high-pitch spiral acquisition mode available in second-generation dual-source CT,
effective doses delivered at CTCA as calculated using the DLP may be reduced to 1 mSv and less [2]. We found a dose reduction of 89% for high-pitch CTCA compared to retrospectively ECG-gated CTCA using the same scan range and tube voltage. However, highpitch protocols are only applicable in patients with low and stable heart rates. In patients with high and irregular heart rates, a narrow diastolic imaging window regarding the RR-interval may not yield diagnostic image quality of the coronary arteries and tube current modulation may need adjustment accordingly [18]. Therefore, in the present study, a full tube current window of 30–80% of the RRinterval was chosen for retrospectively ECG-gated spiral scanning protocols, an acquisition mode that is clinically applied in patients with high and irregular heart rates [14]. For the prospectively ECGtriggered scanning protocols the window was reduced to 60–80% of the RR-interval, according to 70% ± 10% padding, applied in patients with low heart rates yet minor heart rate variations [15]. Radiation doses using these protocols were significantly higher than for the high-pitch protocols at the same scan length. With reported potential dose reductions of up to 86% also for prospectively ECG-triggered sequential acquisition with a narrow acquisition window at 70% of the RR-interval compared to retrospective ECG-gating [19], the importance of adapting the protocol to the heart rate and body mass index of the patient, i. e. using a narrow window of ECG-pulsing at low heart rates and a lowering tube voltage in slim patients, cannot be emphasized enough. Furthermore, the DLP and therefore consecutively the estimated effective dose increases in a linear fashion with increasing scan range. Due to averaging effects, this is also reflected in our results of effective dose calculations according to ICRP guidelines from doses measured using the different scan ranges applied at HP-CTCA which were proportionately higher at the 15 cm scan range compared to the 13.5 cm scan range. Thus, radiation dose can be saved by optimizing the scan range. Image noise may become a limiting factor of image quality when tube voltage and/or tube current are reduced, or if the pitch is increased in order to lower radiation exposure. A tube voltage reduction from 120 kV to 100 kV resulted in a significantly lower radiation dose at the cost of higher image noise. Similarly, the high pitch value used in HP acquisitions significantly increased noise as compared to SEQ and RETRO protocols. However, according to the “as-low-as-reasonably-achievable” (ALARA) principle, image noise is acceptable as long as diagnostic image quality is achieved. All protocols we applied are reported to yield excellent image quality in clinical studies, enabling a high diagnostic accuracy for the assessment of the coronary arteries [1,2,7]. Furthermore, HP acquisition at 120 kV was associated with significantly lower radiation compared to SEQ acquisition at 100 kV while yielding comparable image noise levels. Effective doses as calculated according to ICRP guidelines were lower at high-pitch CTCA than those at CCA with exposure times determined to simulate an optimal, uncomplicated diagnostic examination (1.8–3.9 mSv vs. 4.6–4.8 mSv), however the differences in organ doses were not significant. We chose exposure times at CCA corresponding to an examination performed under optimal conditions for reasons of comparability to HP-CTCA which also requires optimal conditions (i.e. a low and regular heart rate). In a recent head-to-head clinical study however, effective doses were shown to be approximately 4-fold higher at diagnostic CCA with a mean estimated dose of 8.5 mSv compared to 2.1 mSv at prospectively ECG-triggered sequential CTCA using a most narrow window setting at 75% of the RR-interval [20]. Comparisons of reported effective doses for CTCA are problematic, as different methods of calculation may yield substantially different results. For estimations using the DLP, different conversion coefficients of 0.017 mSv/[mGy cm] or 0.014 mSv/[mGy cm] are used, according to the European Guidelines on Quality Criteria for
R. Goetti et al. / European Journal of Radiology 81 (2012) 773–779
Computed Tomography published in 2000 [8] and 2004 [21]. It has previously been reported that effective doses for CTCA are lower when estimated from the DLP than those calculated according to the ICRP 60 guidelines [11]. Furthermore, effective doses at CTCA, as calculated according to the ICRP 103 guidelines, were shown to be up to 36% higher than those calculated according to the ICRP 60 guidelines [22]. This corresponds well with our findings and is mainly attributable to the increased tissue weighting factor of the breast (0.12 vs. 0.05) in the ICRP 103 guidelines [13]. Similarly to our results, effective doses calculated according to ICRP 103 guidelines are reported to be up to 100% higher than those estimated using the DLP and a conversion coefficient of 0.017 mSv/[mGy cm] [3]. Nevertheless, the DLP is sufficient for comparisons of radiation doses between different CTCA protocols, since effective doses calculated according to both IRCP 60 and IRCP 103 guidelines show an excellent correlation with the DLP. Our results suggest that effective doses as calculated according to ICRP 60 and 103 guidelines can be estimated using the DLP and conversion coefficients of k = 0.028 mSv/[mGy cm] and k = 0.034 mSv/[mGy cm], respectively. However these conversion coefficients cannot be directly transferred to other scanners, even of the same model. At CCA effective doses were also higher when calculated according to IRCP 60 and 103 guidelines as compared to the estimate derived from the DAP. These differences however were lower than those for CTCA, most probably because breast doses were lower at CCA compared to CTCA due to the predominance of posterior–anterior and lateral projections at CCA. At CCA the highest effective doses were delivered to the upper abdominal organs (stomach and liver) rather than the breast. 4.1. Limitations A major limitation of our study is that only a male Alderson phantom was used and we did not have a female breast phantom available, therefore two TLDs 1 cm below the skin surface at the level of the breast were used to determine breast dose in each experiment. The organ doses we measured at the breast therefore might not accurately reflect the doses delivered to the female breast. However, the ICRP tissue weighting factors are sex-averaged values for all organs and tissues, including the male and female breast. Further, the used phantom simulates a person of 175 cm height and 73.5 kg weight and does not allow conclusions regarding variations in effective dose levels between larger and smaller patients. Furthermore, all scans were performed on one secondgeneration dual-source CT scanner and our results may not be directly transferred to other scanners, even of the same model. Finally, quantitative image quality assessment was limited to the evaluation of image noise, as the phantom cannot simulate moving coronary arteries and does not allow contrast material administration. 5. Conclusions Dose reduction strategies implemented in CTCA protocols, such as ECG-synchronized tube current modulation, adaption of tube voltage and the use of sequential and high-pitch scanning with prospective ECG-triggering lead to a substantial reduction of radiation doses delivered to the patient. However, caution is warranted in the interpretation of reported effective doses, as significant differences can be found depending on the method of calculation. We found significantly higher effective dose values calculated accord-
779
ing to ICRP guidelines than estimated by calculations based on the DLP and a conversion coefficient of k = 0.017 mSv/[mGy cm], however with an excellent correlation between DLP and ICRP calculations. Our results demonstrate, that effective doses calculated according to ICRP guidelines from experimentally determined doses are significantly lower at HP-CTCA compared to conventional CTCA protocols and are comparable to those at uncomplicated diagnostic CCA. References [1] Johnson TR, Nikolaou K, Busch S, et al. Diagnostic accuracy of dual-source computed tomography in the diagnosis of coronary artery disease. Invest Radiol 2007;42(10):684–91. [2] Leschka S, Stolzmann P, Desbiolles L, et al. Diagnostic accuracy of high-pitch dual-source CT for the assessment of coronary stenoses: first experience. Eur Radiol 2009;19(12):2896–903. [3] Einstein AJ, Elliston CD, Arai AE, et al. Radiation dose from single-heartbeat coronary CT angiography performed with a 320-detector row volume scanner. Radiology 2010;254(3):698–706. [4] Hausleiter J, Meyer T, Hermann F, et al. Estimated radiation dose associated with cardiac CT angiography. JAMA 2009;301(5):500–7. [5] Ketelsen D, Thomas C, Werner M, et al. Dual-source computed tomography: estimation of radiation exposure of ECG-gated and ECG-triggered coronary angiography. Eur J Radiol 2010;73(2):274–9. [6] Abada HT, Larchez C, Daoud B, Sigal-Cinqualbre A, Paul JF. MDCT of the coronary arteries: feasibility of low-dose CT with ECG-pulsed tube current modulation to reduce radiation dose. AJR Am J Roentgenol 2006;186(6 Suppl. 2): S387–90. [7] Stolzmann P, Goetti R, Baumueller S, et al. Prospective and retrospective ECGgating for CT coronary angiography perform similarly accurate at low heart rates. Eur J Radiol 2010. Jan 14. [Epub ahead of print]. [8] Menzel H, Schibilla H, Teunen D. European guidelines on quality criteria for computed tomography, Publication no. EUR 16262 EN. Luxembourg: European Commission; 2000. [9] Kalender WA, Schmidt B, Zankl M, Schmidt M. A PC program for estimating organ dose and effective dose values in computed tomography. Eur Radiol 1999;9(3):555–62. [10] Stamm G, Nagel HD. CT-expo—a novel program for dose evaluation in CT. Rofo 2002;174(12):1570–6. [11] Hunold P, Vogt FM, Schmermund A, et al. Radiation exposure during cardiac CT: effective doses at multi-detector row CT and electron-beam CT. Radiology 2003;226(1):145–52. [12] 1990 Recommendations of the International Commission on Radiological Protection. Ann ICRP 1991;21(1–3):1–201. [13] The 2007 Recommendations of the International Commission on Radiological Protection. ICRP publication 103. Ann ICRP 2007;37(2–4):1–332. [14] Arnoldi E, Johnson TR, Rist C, et al. Adequate image quality with reduced radiation dose in prospectively triggered coronary CTA compared with retrospective techniques. Eur Radiol 2009;19(9):2147–55. [15] Earls JP, Berman EL, Urban BA, et al. Prospectively gated transverse coronary CT angiography versus retrospectively gated helical technique: improved image quality and reduced radiation dose. Radiology 2008;246(3):742–53. [16] Lobotessi H, Karoussou A, Neofotistou V, Louisi A, Tsapaki V. Effective dose to a patient undergoing coronary angiography. Radiat Prot Dosimetry 2001;94(1–2):173–6. [17] Einstein AJ, Moser KW, Thompson RC, Cerqueira MD, Henzlova MJ. Radiation dose to patients from cardiac diagnostic imaging. Circulation 2007;116(11):1290–305. [18] Weustink AC, Neefjes LA, Kyrzopoulos S, et al. Impact of heart rate frequency and variability on radiation exposure, image quality, and diagnostic performance in dual-source spiral CT coronary angiography. Radiology 2009;253(3):672–80. [19] Alkadhi H, Stolzmann P, Scheffel H, et al. Radiation dose of cardiac dual-source CT: the effect of tailoring the protocol to patient-specific parameters. Eur J Radiol 2008;68(3):385–91. [20] Herzog BA, Wyss CA, Husmann L, et al. First head-to-head comparison of effective radiation dose from low-dose 64-slice CT with prospective ECG-triggering versus invasive coronary angiography. Heart 2009;95(20):1656–61. [21] Shrimpton P. Assessment of patient dose in CT. In: EUR. European guidelines for multislice computed tomography funded by the European Commission 2004. Luxembourg: European Commission; 2004. [22] Christner JA, Kofler JM, McCollough CH. Estimating effective dose for CT using dose-length product compared with using organ doses: consequences of adopting international commission on radiological protection publication 103 or dual-energy scanning. AJR Am J Roentgenol 2010;194(4):881–9.
Contents lists available at ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Radiation doses from phantom measurements at high-pitch dual-source computed tomography coronary angiography Robert Goetti a,∗ , Sebastian Leschka b , Markus Boschung c , Sabine Mayer c , Christophe Wyss d , Paul Stolzmann a,e , Thomas Frauenfelder a a
Department of Diagnostic Radiology, University Hospital Zurich, Raemistrasse 100, CH-8091 Zurich, Switzerland Institute of Radiology, Kantonsspital St. Gallen, Rorschacherstrasse 95, CH-9007 St. Gallen, Switzerland c Division for Radiation Safety and Security, Paul Scherrer Institut, CH-5232 Villigen, Switzerland d Division of Cardiology, University Hospital Zurich, Raemistrasse 100, CH-8091 Zurich, Switzerland e Cardiac MR PET CT Program, Massachusetts General Hospital and Harvard Medical School, 165 Cambridge St, Boston, MA 02114, USA b
a r t i c l e
i n f o
Article history: Received 14 September 2010 Received in revised form 6 January 2011 Accepted 17 January 2011 Keywords: Radiation dose High-pitch Computed tomography Coronary angiography Anthropomorphic phantom
a b s t r a c t Objective: To compare radiation doses delivered at prospectively ECG-triggered sequential- (SEQ), retrospectively ECG-gated spiral- (RETRO) and prospectively ECG-triggered high-pitch spiral- (HP) computed tomography coronary angiography (CTCA) protocols, as well as catheter coronary angiography (CCA) using an anthropomorphic phantom. Materials and methods: An anthropomorphic Alderson phantom equipped with 50 thermoluminescent dosimeters (TLDs) was scanned using different CTCA protocols and an uncomplicated diagnostic CCA examination was simulated. Absorbed doses were experimentally determined and effective doses calculated using the dose-length product (DLP) for CTCA and the dose-area product (DAP) for CCA, as well as according to International Commission on Radiation Protection (ICRP) publications 60 and 103. Results: Effective organ doses were significantly lower for HP protocols (100 kV: 0.17 ± 0.26 mSv; 120 kV: 0.26 ± 0.39 mSv) compared to SEQ protocols (100 kV: 0.50 ± 0.79 mSv; 120 kV: 0.90 ± 1.41 mSv; each p < 0.05) and compared to RETRO protocols (100 kV: 1.59 ± 2.12 mSv; 120 kV: 2.75 ± 3.50 mSv; each p < 0.05). Effective organ doses at HP-CTCA tended to be lower than at CCA (0.37 ± 0.40 mSv), however this was not statistically significant (p = 0.13). Effective doses calculated according to ICRP guidelines could be estimated using the DLP and a conversion coefficient of k = 0.034 mSv/[mGy cm] (ICRP103) or k = 0.028 mSv/[mGy cm] (ICRP60), respectively. HP-CTCA led to a dose reduction of 89% compared to RETRO-CTCA, regardless of the calculation method used. Conclusions: Radiation doses as determined by phantom measurements are significantly lower at HPCTCA compared to SEQ-CTCA and RETRO-CTCA and comparable to uncomplicated diagnostic CCA. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Computed tomography coronary angiography (CTCA) has emerged as a robust non-invasive alternative to catheter coronary angiography (CCA) for the diagnosis of coronary artery disease (CAD) [1,2]. However with the introduction of 16- and 64-slice CT systems, rising radiation doses – due to the necessity of higher tube current at higher spatial resolution [3] – have become a concern with reported doses of up to 30 mSv [4] when applying retrospectively ECG-gated acquisition protocols at a low pitch of 0.2–0.5. This concern has led to the development of various strategies to reduce radiation exposure to the patient, including ECG-based tube current modulation [5], tube voltage reduction in slim patients [6],
∗ Corresponding author. Tel.: +41 44 255 11 11; fax: +41 44 255 44 43. E-mail address: [email protected] (R. Goetti). 0720-048X/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2011.01.068
use of sequential scanning [5,7], volume scanning in 320-detector single-source CT [3], and the most recently introduced high-pitch spiral scanning in dual-source CT [2]. Patient radiation dose estimates at CTCA are usually derived from scanner protocol data and a method proposed by the European Working Group for Guidelines on Quality Criteria for CT [8] by multiplication of the dose-length-product (DLP) with a region-specific conversion coefficient. Furthermore, commercially available software applications allow more accurate and individualized estimates of effective doses delivered to different organs by specification of patient parameters such as gender, size, exact extent of scanned body region as well as scanning parameters such as tube voltage, tube current, slice collimation and pitch [9,10]. However, these methods do not actually measure radiation doses delivered to different organs. It has been reported by various authors [3,11] that effective doses estimated using the DLP may be substantially lower than those calculated according to the guide-
774
R. Goetti et al. / European Journal of Radiology 81 (2012) 773–779
Fig. 1. Photograph of the Alderson phantom used for the study (left) with corresponding CT scanogram (right) showing the embedded skeleton and different attenuation of lungs and soft tissue. The box within the CT scanogram image represents the 13.5 cm scan range.
lines of the International Commission on Radiation Protection (ICRP), and vary between calculations according to ICRP publications 60 [12] and 103 [13], due to different tissue weighting factors. Whereas effective doses can be estimated in patient studies using the DLP and a conversion coefficient, effective dose calculations according to ICRP publication 60 and 103 guidelines are obtained by phantom measurements. Such measurements have been reported in various studies for CTCA using retrospectively ECG-gated spiral-, prospectively ECG-triggered sequential and volume acquisition protocols [3,5,11]. To the best of our knowledge, there is however no study in which effective doses at CTCA performed in the high-pitch spiral acquisition mode were calculated from experimentally determined doses and compared to those of conventional CTCA protocols and catheter coronary angiography. The purpose of this study therefore was measure and compare effective doses delivered at CTCA in conventional (i.e. retrospectively ECG-gated spiral acquisition and prospectively ECG-triggered sequential acquisition) and prospectively ECG-triggered high-pitch spiral acquisition protocols, as well as catheter coronary angiography using an anthropomorphic Alderson phantom.
2. Materials and methods 2.1. Dosimetry An anthropomorphic male Alderson phantom was used for all dose measurements (Fig. 1). This phantom consists of a head, neck, and torso, without arms and legs, simulating a male of 175 cm height and 73.5 kg weight. It includes a human skeleton embedded in isocyanide rubber mass (specific density: 0.985 g/cm3 , mean atomic number 7.3) simulating soft tissue, and contains preformed lung substitutes consisting of microcellular foam (specific density: 0.32 g/cm3 , mean atomic number 7.3). It is formed of 35 slices of 2.5 cm thickness, each slice containing a grid of multiple boreholes (spaced by 3 cm from each other) in which thermoluminescent dosimeters (TLDs) can be placed.
Lithium fluoride TLDs (TLD-700; Thermo Fisher Scientific Inc, Waltham, Massachusetts) were used in this study. The dimensions of these dosimeters are 3.2 × 3.2 × 0.9 mm3 . All evaluations of the TLDs were performed with a commercially available reader (Harshaw 5500, Thermo Fisher Scientific Inc, Waltham, Massachusetts). For stabilization, the TLDs were put in an oven for 1 h at a temperature of 90 ◦ C prior to the evaluation. After each evaluation the TLDs were annealed at 400 ◦ C for 1 h. The TLD signal (in nC) was corrected for the individual sensitivity of the TLD and multiplied by a calibration factor to obtain absorbed dose in water (in mGy). For each examination 50 TLDs were positioned inside the phantom. Ten TLDs were used as blanks. A calibration factor was determined in parallel for each examination by irradiating 10 TLDs free-in-air in a reference field of a 137 Cs source. The air kerma at the calibration position was measured with an ionization chamber and converted into absorbed dose in water. Typical calibration factors obtained were 0.4 mGy/nC. The overall uncertainty of the measured doses is about 9%, including the uncertainty of the individual sensitivity (7%) and the uncertainty of the calibration factor (5%). The 50 TLDs used at each examination were positioned at the location of different organs as previously described [11]: bone marrow, esophagus, thyroid gland, lungs, breasts, skin, stomach, liver, colon, ovaries, bladder and testicles. The numbers and locations of the TLDs in the respective slices of the Alderson phantom and according organs are listed in Table 1. Effective doses were calculated from the TLD measurements according to the guidelines published in the International Commission on Radiological Protection (ICRP) Publications 60 and 103 [12,13] by multiplying the mean experimentally determined dose of each organ with the corresponding tissue weighting factors published in ICRP publications 60 and 103. Organ doses of tissues with individual tissue weighting factors in both ICRP 60 and 103 were determined by calculating the mean of at least two (range, 2–12, see Table 1) measurements of TLDs positioned at corresponding locations of the phantom. Doses of remainder tissues were averaged from measurements of adjacent organs. As the TLDs were calibrated in absorbed dose in water, the organ doses were corrected with the organ specific mass energy-absorption coefficient. The uncertainty
R. Goetti et al. / European Journal of Radiology 81 (2012) 773–779 Table 1 Location and number of thermoluminescent dosimeters (TLDs) within the slices of the Alderson phantom. Phantom slice number
Number of TLDs (n = 50)
Organs
3 9 13 14 17 18 21 26 28 29 31 32 34
2 4 (2 + 2)a 6 (2 + 4) 2 6 (2 + 4) 6 (2 + 4) 6 (2 + 2 + 2) 2 4 (2 + 2) 2 4 2 4
Skull (bone marrow) Esophagus, thyroid gland Ribs (bone marrow), lungs Breasts Ribs (bone marrow), lungs Spine (bone marrow), skin Esophagus, stomach, liver Colon Pelvis (bone marrow), colon Sacrum (bone marrow) Ovaries Bladder Testicles
a Numbers in brackets indicate the numbers of TLDs for each organ in the respective slice.
of the effective dose introduced by this correction is in the order of 20%. 2.2. Computed tomography coronary angiography Eight different CTCA protocols were evaluated (Table 2): (a) two retrospectively ECG-gated spiral (RETRO) acquisition protocols (at 100 kV and 120 kV tube voltage, scan range 15 cm) (b) two prospectively ECG-triggered sequential “step-and-shoot” (SEQ) acquisition protocols (at 100 kV and 120 kV tube voltage, scan range 13.5 cm) (c) four prospectively ECG-triggered high-pitch spiral (HP) acquisition protocols (at 100 kV and 120 kV tube voltage; 13.5 cm and 15 cm scan range). All scans were performed on a second-generation dual-source CT system (Somatom Definition Flash, Siemens Healthcare, Forchheim, Germany). Tube current-time product was set at 320 mAs for all scans; tube voltages were 100 kV or 120 kV; pitch was 0.23 for retrospectively ECG-gated spiral scans and 3.2 for prospectively ECG-triggered spiral scans. The scan ranges were 13.5 cm or 15 cm, approximately corresponding to the 50th and 75th percentiles of scan ranges of CTCA reported in a recent multicenter trial [4]. The 15 cm scan range for RETRO acquisitions was chosen to begin at a level 2 cm below the tracheal bifurcation and end at the apex of the heart of the anthropomorphic Alderson phantom. For SEQ acquisitions, scan ranges can only be set in fixed intervals. Thus, four blocks of sequential detector coverage were chosen, corresponding to a scan range of 13.5 cm, with the same center of the scan range as in RETRO acquisitions. To assure comparability of dose values, HP acquisitions were performed at both 13.5 cm and 15 cm
775
scan ranges and all comparisons were made only between acquisitions performed with the same scan range. A heart rate of 60 beats per minute was simulated by the scanner software for all scans and full tube current was applied from 30 to 80% of the RR-interval for retrospectively ECG-gated spiral scans and from 60 to 80% of the RR-interval for prospectively ECG-triggered sequential scans to simulate clinically established protocols [14,15]. For prospectively triggered high-pitch scans, 60% of the RR-interval was set to trigger the scan and full tube current was applied during the entire scan. To achieve a sufficient exposure of the TLDs for accurate measurement of absorbed doses, high-pitch CTCA scans were performed ten times and all other CTCA scans were performed five times. The resulting doses were divided by ten or five, respectively, to calculate the reported single-scan experimentally determined doses. Effective doses at CTCA were estimated using the method proposed by the European Working Group for Guidelines on Quality Criteria for CT [8] using the dose-length product (DLP) and a conversion coefficient for the chest (k = 0.017 mSv/[mGy cm]) and were calculated from the experimentally determined doses according to the guidelines in ICRP publications 60 and 103 as described above. To compare image quality between the different protocols, image noise was measured as follows: for all datasets (HP, SEQ and RETRO, each at 100 kV and 120 kV) images were reconstructed using standard cardiac imaging parameters (slice thickness 0.75 mm, increment 0.5 mm, convolution kernel B26f). Regions of interest (ROIs) of 6 cm2 – avoiding boreholes – were drawn at five different locations (ascending aorta, aortic root, basal heart, midventricular heart and apical heart) in each image set. Noise was defined as the standard deviation from the attenuation measurements in these ROIs. All measurements were performed by two radiologists blinded to the acquisition mode.
2.3. Catheter coronary angiography For the simulation of a diagnostic CCA examination, a biplane angiographic system was used (Integris Allura 9, Philips Medical Systems, Best, The Netherlands). Exposure times for an uncomplicated diagnostic examination were determined and the examination was simulated by an experienced interventional cardiologist (with seven years of experience in interventional cardiology, anonymized) as follows: firstly, 10 s anterior–posterior (AP) fluoroscopy to simulate the intubation of the left coronary artery (LCA) and 10 s 60◦ left anterior oblique (LAO 60) fluoroscopy simulating the intubation of the right coronary artery (RCA). Then, the simulation of LCA angiography was performed with three exposures in six orientations and RCA angiography with two exposures in four orientations. Each exposure consisted of 1 s fluoroscopy and 2 s cine-imaging. Detailed exposure parameters of the examination are provided in Table 3. In order to achieve a sufficient exposure of the TLDs for accurate measurement of effective doses, the entire examination was performed five times and the total dose divided
Table 2 Computed tomography coronary angiography protocols. Protocol namea
ECG-synchronisation
Acquisition
Tube voltage (kV)
Tube current (mAs)
ECG-pulsing (% of RR-interval)
Pitch
CTDIvol (mGy)
Scan length (mm)
HP100S HP120S HP100L HP120L SEQ100 SEQ120 RETRO100 RETRO120
Prospective Prospective Prospective Prospective Prospective Prospective Retrospective Retrospective
Spiral Spiral Spiral Spiral Sequential Sequential Spiral Spiral
100 120 100 120 100 120 100 120
320 320 320 320 320 320 320 320
– – – – 60–80% 60–80% 30–80% 30–80%
3.2 3.2 3.2 3.2 – – 0.23 0.23
3.10 5.18 3.10 5.18 14.14 23.65 34.53 60.81
135 135 150 150 135 135 150 150
a HP: high-pitch; SEQ: sequential; RETRO: retrospectively ECG-gated; 100: 100 kV tube voltage; 120: 120 kV tube voltage; S: short scan range (135 mm); L: long scan range (150 mm).
776
R. Goetti et al. / European Journal of Radiology 81 (2012) 773–779
Table 3 Orientations, tube voltages and tube currents for the three left coronary artery exposures and two right coronary artery exposures of simulated catheter coronary angiography. Exposure
Orientations
Tube voltage (kV)
Tube current (mAs)
LCA 1
AP LAO90
67 73
439 751
LCA 2
RAO30 CAUD20 LAO40 CRAN20
69
519
73
727
RAO15 CRAN30 LAO50 CAUD20
72
681
74
782
RCA 1
RAO30 LAO30
69 75
547 841
RCA 2
LAO15 CRAN30 LAO90
68
495
74
782
LCA 3
LCA 1–3: left coronary artery exposures; AP: anterior–posterior; LAO90: left anterior oblique, 90◦ angle; RAO30: right anterior oblique, 30◦ angle; CAUD20: 20◦ caudal angulation, CRAN20: 20◦ cranial angulation; RCA 1–2: right coronary artery exposures.
by five. Effective doses at CCA were estimated by multiplying the dose-area product (DAP) of the CCA examination with a conversion coefficient for the chest (k = 0.22 mSv/[Gy cm2 ]) [16] and were calculated from the TLD measurements according to the guidelines in ICRP publications 60 and 103 as described above. 2.4. Statistical analyses All statistical analyses were performed using commercially available software (PASW Statistics 18, SPSS Inc., Chicago, IL). Continuous variables are expressed as means ± standard deviations and categorical variables are expressed as frequencies or percentages. Normality testing was performed using the Kolmogorov–Smirnov test. Differences in experimentally determined organ doses between the CTCA and CCA protocols and differences in experimentally determined organ doses between CTCA protocols at 100 kV and 120 kV tube voltage as well as differences between effective doses calculated according to ICRP 60 and 103 guidelines and DLP were assessed using the paired-samples t-test. Correlations between effective doses at CTCA calculated according to ICRP 60 and 103 guidelines and DLP were assessed using Pearson’s correlation coefficient and linear regression analysis. Bland–Altman analysis was used to assess differences in effective doses between calculations according to ICRP and using the DLP (for CTCA) or using the DAP (for CCA). Differences in image noise measurements between the different CTCA protocols were evaluated with the paired samples t-test. For all tests, a two-tailed p < 0.05 was considered to indicate statistical significance. 3. Results 3.1. Computed tomography coronary angiography Table 4 provides an overview of experimentally determined doses (mGy) as well as calculated effective doses (mSv) for the different protocols. Experimentally determined as well as calculated effective doses were distributed normally according to the Kolmogorov–Smirnov test and were therefore compared using paired samples t-tests. The highest effective organ doses (mSv) as calculated according to ICRP103 guidelines from experimentally determined doses at CTCA were delivered to
Fig. 2. Scatterplots demonstrating excellent correlation between effective doses of the eight CTCA protocols calculated according to the guidelines of the International Commission on Radiation Protection (ICRP) publications 60 (triangles) and 103 (circles) and the dose-length product (DLP). Dotted lines indicate 95% confidence intervals.
the breast, followed by the lungs. Effective organ doses were significantly lower for HP protocols (100 kV: 0.17 ± 0.26 mSv; 120 kV: 0.26 ± 0.39 mSv) compared to SEQ protocols (100 kV: 0.50 ± 0.79 mSv; 120 kV: 0.90 ± 1.41 mSv; each p < 0.05) when using the same scan length of 13.5 cm, and compared to RETRO protocols (100 kV: 0.19 ± 0.26 mSv vs. 1.59 ± 2.12 mSv; 120 kV: 0.29 ± 0.40 mSv vs. 2.75 ± 3.50 mSv; each p < 0.05) when using the same scan length of 15 cm. Experimentally determined organ doses using 100 kV protocols were significantly lower than those at 120 kV protocols (each p < 0.05). Effective doses delivered at CTCA as calculated according to the ICRP guidelines showed a wide range from 1.8 mSv to 35.8 mSv depending on the applied protocol and the different tissue weighting factors for calculations according to ICRP 60 and ICRP 103. Bland–Altman analysis revealed a higher difference between effective doses as estimated using the DLP and effective dose as calculated according to ICRP 103 guidelines (5.4 ± 6.0 mSv) than between effective doses estimated using the DLP and calculated according to ICRP 60 (3.4 ± 3.9 mSv). Effective doses as calculated according to ICRP 103 guidelines were 23.5 ± 3.1% higher than those calculated according to the ICRP 60 guidelines, and 103.4 ± 8.2% higher than those estimated using the DLP and a conversion coefficient of k = 0.017 mSv/[mGy cm] (each p < 0.05). There was however an excellent correlation between effective doses calculated according to ICRP 60 and 103 guidelines and DLP (R2 = 0.999, R2 = 1; each p < 0.001) for the different CTCA protocols (Fig. 2). Linear regression analysis revealed that effective doses according to ICRP calculations could be estimated using the DLP provided in the patient protocol of this specific scanner and a conversion coefficient of k = 0.034 mSv/[mGy cm] (ICRP103) or k = 0.028 mSv/[mGy cm] (ICRP60), respectively. Using the same scan range of 15 cm and a tube current of 120 kV, applying the prospectively ECG-triggered high-pitch spiral protocol led to a dose reduction of 89% compared to the retrospectively ECG-gated protocol, regardless of the calculation method used. There was no significant difference in image noise between
Table 4 Experimentally determined and effective doses at CTCA and CCA. Protocol a
Organ
HP100S
HP100L
HP120S
HP120L
Bone marrow Esophagus Thyroid Lungs Breast Skin Stomach Liver Colon Ovaries Testicles Bladder
1.56 3.11 0.48 5.35 6.73 4.78 1.91 1.77 0.10 0.02 0.01 0.02
1.71 2.68 0.47 4.97 6.82 4.53 3.11 2.83 0.13 0.03 0.02 0.02
2.79 5.19 1.07 9.22 9.09 8.34 2.55 2.19 0.17 0.04 0.03 0.03
3.01 4.54 0.83 8.71 9.86 7.51 4.53 5.11 0.25 0.05 0.03 0.04
61 –
DLP (mGy cm) DAP (mGy cm2 ) Effective dose (mSv)
From DLPb From DAPc ICRP 60 ICRP 103
1.0 – 1.8 2.2
65 – 1.1 – 1.9 2.3
102 – 1.7 – 2.8 3.4
109 – 1.9 – 3.2 3.9
SEQ100 4.83 11.07 1.10 14.84 21.49 18.13 3.84 4.13 0.27 0.07 0.06 0.05 195 – 3.3 – 5.0 6.3
SEQ120
RETRO100
RETRO120
CCA
8.39 19.11 1.98 25.89 38.66 33.79 8.60 7.06 0.57 0.14 0.08 0.10
15.38 29.77 2.49 40.83 57.41 55.91 23.31 27.88 1.33 0.25 0.13 0.17
27.46 51.32 4.77 72.07 87.70 89.93 51.11 42.97 2.79 0.58 0.26 0.38
7.49 8.08 0.13 3.71 1.59 12.60 9.26 22.88 1.05 0.16 0.03 0.09
326 – 5.5 – 9.1 11.7
603 – 10.3 – 16.6 20.6
1043 – 17.7 – 29.5 35.8
– 14,313 – 3.1 4.6 4.8
HP100S: prospectively ECG-triggered high-pitch spiral acquisition at 100 kV with 13.5 cm scan range; HP100L: prospectively ECG-triggered high-pitch spiral acquisition at 100 kV with 15 cm scan range; HP120S: prospectively ECG-triggered high-pitch spiral acquisition at 120 kV with 13.5 cm scan range; HP120L: prospectively ECG-triggered high-pitch spiral acquisition at 120 kV with 15 cm scan range; SEQ100: prospectively ECG-triggered sequential acquisition at 100 kV with 13.5 cm scan range; SEQ120: prospectively ECG-triggered sequential acquisition at 120 kV with 13.5 cm scan range; RETRO100: retrospectively ECG-gated spiral acquisition at 100 kV with 15 cm scan range; RETRO120: retrospectively ECG-gated spiral acquisition at 120 kV with 15 cm scan range. a Only organs with individually specified tissue weighting factors both in ICRP publications 60 and 103 are shown. b Effective dose calculated for CTCA by multiplying the DLP with a conversion coefficient of k = 0.017 mSv/[mGy cm]. c Effective dose calculated for CCA by multiplying the DAP with a conversion coefficient of k = 0.22 mSv/[Gy cm2 ].
R. Goetti et al. / European Journal of Radiology 81 (2012) 773–779
Experimentally determined dose (mGy)
777
778
R. Goetti et al. / European Journal of Radiology 81 (2012) 773–779
the measurements performed by the two readers (15.1 ± 2.2 vs. 15.0 ± 2.3, p = 0.33), indicating good inter-reader agreement. Image noise was significantly lower in 120 kV protocols than in 100 kV protocols (14.0 ± 1.2 vs. 16.2 ± 2.5, p < 0.001). There were no significant differences in image noise between the SEQ and RETRO protocols at 100 kV (14.7 ± 1.6 vs. 14.8 ± 1.6, p = 0.87) and 120 kV (13.4 ± 0.6 vs. 13.7 ± 1.1, p = 0.22), respectively. However, HP protocols (100 kV: 19.0 ± 1.5; 120 kV: 14.8 ± 1.2) yielded significantly higher image noise compared to SEQ and RETRO acquisitions at the same tube voltages (all p < 0.05). Image noise in HP acquisition at 120 kV was comparable to image noise in SEQ acquisition at 100 kV (p = 0.72) whereas experimentally determined doses were significantly lower using the HP protocol at 120 kV compared to the SEQ protocol at 100 kV (3.4 ± 3.6 mGy vs. 6.7 ± 7.7 mGy, p < 0.05). 3.2. Catheter coronary angiography At CCA, the highest effective doses as calculated from experimentally determined doses according to ICRP guidelines were delivered at the upper abdominal organs stomach and liver (Table 4). The effective dose calculated for CCA was 4.6 mSv (ICRP 60) or 4.8 mSv (IRCP 103), which are 48.4–54.8% higher than the effective dose estimated by calculation using the DAP. 3.3. CTCA vs. CCA Effective doses calculated according to ICRP guidelines were lower in the four different HP-CTCA protocols (range, 1.8–3.9 mSv) compared to those at CCA (range, 4.6–4.8 mSv), however the differences in effective organ doses did not reach statistical significance (CCA: 0.37 ± 0.40 mSv; HP-CTCA, 100 kV, 13.5 cm scan range: 0.17 ± 0.26 mSv, p = 0.13; HP-CTCA, 120 kV, 15 cm scan range: 0.29 ± 0.40 mSv, p = 0.57). The highest effective organ doses were delivered to the breast and lungs at CTCA and to the stomach and liver at CCA. 4. Discussion The main findings of our study are: 1. High-pitch CTCA protocols are associated with significantly lower effective doses, than both prospectively ECG-triggered sequential protocols and retrospectively ECG-gated spiral protocols with potential dose reductions of up to 89% and effective doses comparable to those delivered at uncomplicated diagnostic CCA. 2. Effective dose estimates vary substantially depending on the method of calculation for both CTCA and CCA and are significantly different between calculations according to ICRP guidelines and calculations using the DLP and a conversion coefficient of k = 0.017 mSv/[mGy cm] for CTCA, however with an excellent correlation between DLP and ICRP calculations. The wide range of radiation doses at CTCA reported for different protocols and at different centers is an issue of concern pointed out by various authors [4,17] with a reported range of median effective doses varying between 5 mSv and 30 mSv [4]. Our results for SEQ and RETRO acquisition protocols are comparable to those reported in a recent dual-source CTCA phantom study, with effective doses ranging from 2.8 mSv to 32.6 mSv – depending on gender, heart rate acquisition protocol and ECG-pulsing – when taking into account that those measurements were made using a shorter scan range, variable low pitch and different ECG-pulsing windows in a firstgeneration dual-source scanner [5]. Applying the high-pitch spiral acquisition mode available in second-generation dual-source CT,
effective doses delivered at CTCA as calculated using the DLP may be reduced to 1 mSv and less [2]. We found a dose reduction of 89% for high-pitch CTCA compared to retrospectively ECG-gated CTCA using the same scan range and tube voltage. However, highpitch protocols are only applicable in patients with low and stable heart rates. In patients with high and irregular heart rates, a narrow diastolic imaging window regarding the RR-interval may not yield diagnostic image quality of the coronary arteries and tube current modulation may need adjustment accordingly [18]. Therefore, in the present study, a full tube current window of 30–80% of the RRinterval was chosen for retrospectively ECG-gated spiral scanning protocols, an acquisition mode that is clinically applied in patients with high and irregular heart rates [14]. For the prospectively ECGtriggered scanning protocols the window was reduced to 60–80% of the RR-interval, according to 70% ± 10% padding, applied in patients with low heart rates yet minor heart rate variations [15]. Radiation doses using these protocols were significantly higher than for the high-pitch protocols at the same scan length. With reported potential dose reductions of up to 86% also for prospectively ECG-triggered sequential acquisition with a narrow acquisition window at 70% of the RR-interval compared to retrospective ECG-gating [19], the importance of adapting the protocol to the heart rate and body mass index of the patient, i. e. using a narrow window of ECG-pulsing at low heart rates and a lowering tube voltage in slim patients, cannot be emphasized enough. Furthermore, the DLP and therefore consecutively the estimated effective dose increases in a linear fashion with increasing scan range. Due to averaging effects, this is also reflected in our results of effective dose calculations according to ICRP guidelines from doses measured using the different scan ranges applied at HP-CTCA which were proportionately higher at the 15 cm scan range compared to the 13.5 cm scan range. Thus, radiation dose can be saved by optimizing the scan range. Image noise may become a limiting factor of image quality when tube voltage and/or tube current are reduced, or if the pitch is increased in order to lower radiation exposure. A tube voltage reduction from 120 kV to 100 kV resulted in a significantly lower radiation dose at the cost of higher image noise. Similarly, the high pitch value used in HP acquisitions significantly increased noise as compared to SEQ and RETRO protocols. However, according to the “as-low-as-reasonably-achievable” (ALARA) principle, image noise is acceptable as long as diagnostic image quality is achieved. All protocols we applied are reported to yield excellent image quality in clinical studies, enabling a high diagnostic accuracy for the assessment of the coronary arteries [1,2,7]. Furthermore, HP acquisition at 120 kV was associated with significantly lower radiation compared to SEQ acquisition at 100 kV while yielding comparable image noise levels. Effective doses as calculated according to ICRP guidelines were lower at high-pitch CTCA than those at CCA with exposure times determined to simulate an optimal, uncomplicated diagnostic examination (1.8–3.9 mSv vs. 4.6–4.8 mSv), however the differences in organ doses were not significant. We chose exposure times at CCA corresponding to an examination performed under optimal conditions for reasons of comparability to HP-CTCA which also requires optimal conditions (i.e. a low and regular heart rate). In a recent head-to-head clinical study however, effective doses were shown to be approximately 4-fold higher at diagnostic CCA with a mean estimated dose of 8.5 mSv compared to 2.1 mSv at prospectively ECG-triggered sequential CTCA using a most narrow window setting at 75% of the RR-interval [20]. Comparisons of reported effective doses for CTCA are problematic, as different methods of calculation may yield substantially different results. For estimations using the DLP, different conversion coefficients of 0.017 mSv/[mGy cm] or 0.014 mSv/[mGy cm] are used, according to the European Guidelines on Quality Criteria for
R. Goetti et al. / European Journal of Radiology 81 (2012) 773–779
Computed Tomography published in 2000 [8] and 2004 [21]. It has previously been reported that effective doses for CTCA are lower when estimated from the DLP than those calculated according to the ICRP 60 guidelines [11]. Furthermore, effective doses at CTCA, as calculated according to the ICRP 103 guidelines, were shown to be up to 36% higher than those calculated according to the ICRP 60 guidelines [22]. This corresponds well with our findings and is mainly attributable to the increased tissue weighting factor of the breast (0.12 vs. 0.05) in the ICRP 103 guidelines [13]. Similarly to our results, effective doses calculated according to ICRP 103 guidelines are reported to be up to 100% higher than those estimated using the DLP and a conversion coefficient of 0.017 mSv/[mGy cm] [3]. Nevertheless, the DLP is sufficient for comparisons of radiation doses between different CTCA protocols, since effective doses calculated according to both IRCP 60 and IRCP 103 guidelines show an excellent correlation with the DLP. Our results suggest that effective doses as calculated according to ICRP 60 and 103 guidelines can be estimated using the DLP and conversion coefficients of k = 0.028 mSv/[mGy cm] and k = 0.034 mSv/[mGy cm], respectively. However these conversion coefficients cannot be directly transferred to other scanners, even of the same model. At CCA effective doses were also higher when calculated according to IRCP 60 and 103 guidelines as compared to the estimate derived from the DAP. These differences however were lower than those for CTCA, most probably because breast doses were lower at CCA compared to CTCA due to the predominance of posterior–anterior and lateral projections at CCA. At CCA the highest effective doses were delivered to the upper abdominal organs (stomach and liver) rather than the breast. 4.1. Limitations A major limitation of our study is that only a male Alderson phantom was used and we did not have a female breast phantom available, therefore two TLDs 1 cm below the skin surface at the level of the breast were used to determine breast dose in each experiment. The organ doses we measured at the breast therefore might not accurately reflect the doses delivered to the female breast. However, the ICRP tissue weighting factors are sex-averaged values for all organs and tissues, including the male and female breast. Further, the used phantom simulates a person of 175 cm height and 73.5 kg weight and does not allow conclusions regarding variations in effective dose levels between larger and smaller patients. Furthermore, all scans were performed on one secondgeneration dual-source CT scanner and our results may not be directly transferred to other scanners, even of the same model. Finally, quantitative image quality assessment was limited to the evaluation of image noise, as the phantom cannot simulate moving coronary arteries and does not allow contrast material administration. 5. Conclusions Dose reduction strategies implemented in CTCA protocols, such as ECG-synchronized tube current modulation, adaption of tube voltage and the use of sequential and high-pitch scanning with prospective ECG-triggering lead to a substantial reduction of radiation doses delivered to the patient. However, caution is warranted in the interpretation of reported effective doses, as significant differences can be found depending on the method of calculation. We found significantly higher effective dose values calculated accord-
779
ing to ICRP guidelines than estimated by calculations based on the DLP and a conversion coefficient of k = 0.017 mSv/[mGy cm], however with an excellent correlation between DLP and ICRP calculations. Our results demonstrate, that effective doses calculated according to ICRP guidelines from experimentally determined doses are significantly lower at HP-CTCA compared to conventional CTCA protocols and are comparable to those at uncomplicated diagnostic CCA. References [1] Johnson TR, Nikolaou K, Busch S, et al. Diagnostic accuracy of dual-source computed tomography in the diagnosis of coronary artery disease. Invest Radiol 2007;42(10):684–91. [2] Leschka S, Stolzmann P, Desbiolles L, et al. Diagnostic accuracy of high-pitch dual-source CT for the assessment of coronary stenoses: first experience. Eur Radiol 2009;19(12):2896–903. [3] Einstein AJ, Elliston CD, Arai AE, et al. Radiation dose from single-heartbeat coronary CT angiography performed with a 320-detector row volume scanner. Radiology 2010;254(3):698–706. [4] Hausleiter J, Meyer T, Hermann F, et al. Estimated radiation dose associated with cardiac CT angiography. JAMA 2009;301(5):500–7. [5] Ketelsen D, Thomas C, Werner M, et al. Dual-source computed tomography: estimation of radiation exposure of ECG-gated and ECG-triggered coronary angiography. Eur J Radiol 2010;73(2):274–9. [6] Abada HT, Larchez C, Daoud B, Sigal-Cinqualbre A, Paul JF. MDCT of the coronary arteries: feasibility of low-dose CT with ECG-pulsed tube current modulation to reduce radiation dose. AJR Am J Roentgenol 2006;186(6 Suppl. 2): S387–90. [7] Stolzmann P, Goetti R, Baumueller S, et al. Prospective and retrospective ECGgating for CT coronary angiography perform similarly accurate at low heart rates. Eur J Radiol 2010. Jan 14. [Epub ahead of print]. [8] Menzel H, Schibilla H, Teunen D. European guidelines on quality criteria for computed tomography, Publication no. EUR 16262 EN. Luxembourg: European Commission; 2000. [9] Kalender WA, Schmidt B, Zankl M, Schmidt M. A PC program for estimating organ dose and effective dose values in computed tomography. Eur Radiol 1999;9(3):555–62. [10] Stamm G, Nagel HD. CT-expo—a novel program for dose evaluation in CT. Rofo 2002;174(12):1570–6. [11] Hunold P, Vogt FM, Schmermund A, et al. Radiation exposure during cardiac CT: effective doses at multi-detector row CT and electron-beam CT. Radiology 2003;226(1):145–52. [12] 1990 Recommendations of the International Commission on Radiological Protection. Ann ICRP 1991;21(1–3):1–201. [13] The 2007 Recommendations of the International Commission on Radiological Protection. ICRP publication 103. Ann ICRP 2007;37(2–4):1–332. [14] Arnoldi E, Johnson TR, Rist C, et al. Adequate image quality with reduced radiation dose in prospectively triggered coronary CTA compared with retrospective techniques. Eur Radiol 2009;19(9):2147–55. [15] Earls JP, Berman EL, Urban BA, et al. Prospectively gated transverse coronary CT angiography versus retrospectively gated helical technique: improved image quality and reduced radiation dose. Radiology 2008;246(3):742–53. [16] Lobotessi H, Karoussou A, Neofotistou V, Louisi A, Tsapaki V. Effective dose to a patient undergoing coronary angiography. Radiat Prot Dosimetry 2001;94(1–2):173–6. [17] Einstein AJ, Moser KW, Thompson RC, Cerqueira MD, Henzlova MJ. Radiation dose to patients from cardiac diagnostic imaging. Circulation 2007;116(11):1290–305. [18] Weustink AC, Neefjes LA, Kyrzopoulos S, et al. Impact of heart rate frequency and variability on radiation exposure, image quality, and diagnostic performance in dual-source spiral CT coronary angiography. Radiology 2009;253(3):672–80. [19] Alkadhi H, Stolzmann P, Scheffel H, et al. Radiation dose of cardiac dual-source CT: the effect of tailoring the protocol to patient-specific parameters. Eur J Radiol 2008;68(3):385–91. [20] Herzog BA, Wyss CA, Husmann L, et al. First head-to-head comparison of effective radiation dose from low-dose 64-slice CT with prospective ECG-triggering versus invasive coronary angiography. Heart 2009;95(20):1656–61. [21] Shrimpton P. Assessment of patient dose in CT. In: EUR. European guidelines for multislice computed tomography funded by the European Commission 2004. Luxembourg: European Commission; 2004. [22] Christner JA, Kofler JM, McCollough CH. Estimating effective dose for CT using dose-length product compared with using organ doses: consequences of adopting international commission on radiological protection publication 103 or dual-energy scanning. AJR Am J Roentgenol 2010;194(4):881–9.