| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Imaging, Division of Nuclear Medicine; and Department of
Medicine, Division of Cardiology, Cedars-Sinai Medical Center, 8700 Beverly
Blvd., Rm. 1258, Los Angeles, CA 90048.
2 Heart Disease Prevention Program, Division of Cardiology, University of
California, Irvine, CA.
Received March 2, 2004;
accepted after revision December 6, 2004.
Address correspondence to D. S. Berman
(bermand{at}cshs.org).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Sixty-eight patients underwent both MDCT and EBT within 2 months for the quantification of CAC. The images were scored in a blinded fashion and independently by two observers with a minimum of 7 days between the interpretations of images obtained from one scanner type to the other.
RESULTS. Presence versus absence of CAC was discordant by EBT versus MDCT in 6% (n = 4) of the cases by observer 1, with one of these cases also discordant by observer 2. All cases except one (aortic calcium misidentified as CAC) were among those with a mean Agatston score of less than 5 present on EBT but absent on MDCT. EBT and MDCT scores correlated well (r = 0.98-0.99). The relative median variability between EBT and MDCT for the Agatston score was 24% for observer 1 and 27% for observer 2 and was 18% and 14%, respectively, for volume score (average for both observers: 27% for Agatston score and 16% for volume score). Scores were higher for EBT than MDCT in approximately half of the cases, with little systematic difference between the two (median EBT-MDCT difference: Agatston score, -0.55; volume score, 3.4 mm3). The absolute median difference averaged for the two observers was 28.75 for the Agatston score and 15.4 mm3 for the volume score.
CONCLUSION. Differences in CAC measurements using EBT and MDCT are similar to interscan differences in CAC measurements previously reported for EBT or for other MDCT scanners individually.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Scanning was performed on an EBT scanner (C150XP, Imatron) and an MDCT scanner (Somatom Volume Zoom, Siemens Medical Solutions) with a maximum interval of 2 months between examinations. For EBT, we used a protocol of 3-mm slice thickness, 100-msec exposure time, 130 kVp, 630-mA tube current (yielding 63 mAs for 100-msec exposure), prospective ECG triggering at 60% of the R-R interval, either a 300- or 350-mm field of view for reconstruction, and the sharp reconstruction kernel. For MDCT, the protocol was 2.5-mm slice thickness, 120 kVp, 168-mA tube current (yielding 42 mAs for 250-msec exposure), 250-msec exposure time, prospective ECG triggering at 400 msec before the next R wave, and a 350-mm field of view. The entire heart was covered in a single breath-hold for both examinations.
The images from both scanners were stored in an archive for preprocessing before interpretation. A custom software tool written by our laboratory in the Java language (Sun Corporation) was used to code the image file names to blind the observers and scramble the order of presentation.
For interpretation, the images were transferred to a dedicated workstation (NetraMD, ScImage). Using the Calcified Plaque Analysis software tool of the NetraMD workstation, each observer scored all the images from one scanner and then, after a minimum of 7 days, all the images from the second scanner. The information block of the workstation was concealed so that patient-identifying information was hidden from the observers.
There were two observers scoring the data. Observer 1 is an experienced cardiac imaging physician, and observer 2 is an experienced radiologic technologist who routinely scores clinical calcium studies before physician review at our center.
After completion of the blinded interpretation, a second adjudication interpretation was performed by observer 1 with both scans from the same patient simultaneously available to that observer. The purpose of this adjudication was to ensure that every coronary lesion was scored on both scans (of the same patient) and to reclassify corresponding calcifications on each scan (of the same patient) as either both aortic calcium (i.e., not included in calcium score) or both coronary calcium (i.e., both included in calcium score). The purpose of this second interpretation was to determine the agreement of the scanners without the influence of intraobserver variability.
Statistical analysis involved assessment of concordance between the two
techniques with respect to whether scan was positive (nonzero vs zero score)
and to calcium score category (0-9, 10-99, 100-399, and 
400), and kappa
statistics for agreement were determined. Pearson correlation coefficients
among scores obtained by the two techniques were also determined.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
for perfect agreement within the same
classification = 0.558). Results were similar for observer 2 (not shown). EBT
and MDCT scores for observer 1 correlated well (r = 0.96-0.97), as
shown in Figures 1 and
2, which display results for
observer 1 for Agatston score and volume score, respectively. Correlations
were similar (0.97-0.98) for observer 2 (data not shown).
|
|
|
|
The median relative variability between EBT and MDCT for Agatston score was 24% for observer 1 and 27% for observer 2 and for volume score was 18% and 14% for the respective observers (average for both observers: 27% for Agatston score and 16% for volume score) (Figs. 3 and 4), and variability decreased directly with increasing extent of Agatston score. Scores were higher for EBT than for MDCT in approximately half of cases, with little systematic difference between the two (median EBT-MDCT difference, Agatston score -0.55; volume score, 3.4 mm3). Relative variability was 100% or greater for scores less than 10, whereas for scores of 400 or greater, it was 13% for Agatston score and 9% for volume score.
|
|
|
|
For the adjudication, the following results were obtained. When EBT and MDCT scans were reevaluated unblinded and side-by-side by observer 1, changes in total scores and volumes occurred for three cases scanned on EBT, three different cases scanned on MDCT, and one additional case scanned on both EBT and MDCT. This resulted in interscan variability for either Agatston score (22.3%) or volume score (18.0%) that was essentially unchanged from the unadjudicated results.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our findings are similar to those of previous reports. Becker et al. [1] described agreement between MDCT and EBT for a similar protocol with the principal difference being the peak kilovoltage setting. As in our study, the results of the study conducted by Becker et al. showed lower variability between MDCT and EBT for the volume score than for the Agatston score and showed high correlations among scores and volumes obtained for both EBT and MDCT. A more recent study by Stanford et al. [2] compared 4-MDCT with EBT for CAC scoring. In that article, a different scanner (Aquilion, Toshiba), different peak kilovoltage setting (135 kVp) for both EBT and MDCT, different acquisition time (320 msec), different slice thickness (3 mm), and different trigger point (80% of R-R interval) were used compared with our work. However, their conclusions were similar.
We consider peak kilovoltage to be an important parameter to explore because calcium attenuation depends on the X-ray beam energy spectrum. Thus, we expect that the agreement for CAC between the gold standard EBT and other types of scanners to, in part, depend on the peak kilovoltage setting. The optimal setting is still open to exploration in our opinion. Our reported protocol is based on sequential prospectively triggered acquisition. There have been studies by other groups using helical (including retrospectively gated) acquisitions [6-10] and a recommendation by Achenbach et al. [11] for helical as a standard for coronary calcium scoring. There are some arguments possible for this mode of use, although the question of the additional dose relative to sequential acquisition needs to be carefully considered in terms of risk and benefit.
Because our objective was to compare the ability of a commonly used MDCT protocol with that of EBT in the assessment of CAC, it was beyond the scope of the current study to examine whether gating techniques, heart rate, slice thickness, or different dose estimates affect reproducibility. Also, it is possible that higher heart rates may produce more motion artifact; however, it was not the purpose of our study to compare reproducibility over a wide range of heart rates. A report by Hong et al. [12] states that interacquisition variability at MDCT is significantly less at lower heart rates. Finally, slice thickness is an important parameter that can affect the observed attenuation of calcium depending on whether partial volume effects are an important driver in Agatston scoring. A report by Vliegenthart et al. [13] describes improvements by using 1.5-mm slices versus 3.0-mm slices at EBT. Moreover, Wang et al. [14] reported lower interscan variability from 6-mm slice thicknesses (14%) as compared with 3-mm slice thicknesses (29%). In our work, EBT was set at 3-mm slices and MDCT was at 2.5-mm slices. Because we obtained only one scan on each technique per patient, we cannot assess variability that is purely attributable to slice thickness.
We used triggering at 400 msec before the next R wave at MDCT (and 60% of the R-R interval at EBT). Since this was a prospective technique, it was not possible to reconstruct the same patient data at other phases because acquisitions were not continuous as in the case of retrospective helical MDCT. Thus, the question of how well other phases would do is somewhat open at MDCT, but probably would depend on the specific patient. It is likely that an algorithm to determine this automatically from a patient's ECG signal at acquisition would be a useful approach. In fact, a report by Lu et al. [15] describes improvements in reproducibility of CAC measurements at EBT by use of optimal ECG triggering.
Although we report on the Agatston and volume scores, it is possible that the calcium mass may be more reproducible and would agree better between scanners. However, we did not have mass scores available so were unable to assess this. Moreover, the Agatston score is currently more accepted than the mass score and we wished to evaluate performance with respect to this well-known measure that is likely to remain in common use.
The question of relative radiation dose between the two techniques is controversial. Although performing dosimetry on each patient would be outside the scope of this work, similar protocols for calcium scanning generally result in dosage estimates of 0.7 mSv for EBT and 1.1 mSv for MDCT (using a Volume Zoom scanner, Siemens Medical Solutions). We chose the prospective method to minimize radiation exposure to the patient. Our study did not seek to compare sequential prospective gating techniques with other methods such as retrospective gating.
In conclusion, our data confirm earlier reports of a close concordance of CAC estimates between MDCT and EBT. Reproducibility obtained from scanning the same patients on these two different technologies appears similar to that obtained from scanning with individual technologies. This may also have implications in prospective studies involving serial scanning of patients when the same technology may not be available on rescanning. Our report suggests that scanner technology is unlikely to have a marked difference in scores obtained.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  K. A. Barraclough, L. A. Stevens, L. Er, D. Rosenbaum, J. Brown, P. Tiwari, and A. Levin Coronary artery calcification scores in patients with chronic kidney disease prior to dialysis: reliability as a trial outcome measure Nephrol. Dial. Transplant., October 1, 2008; 23(10): 3199 - 3205. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Dey, T. Callister, P. Slomka, F. Aboul-Enein, H. Nishina, X. Kang, H. Gransar, N. D. Wong, R. Miranda-Peats, S. Hayes, et al. Computer-Aided Detection and Evaluation of Lipid-Rich Plaque on Noncontrast Cardiac CT Am. J. Roentgenol., June 1, 2006; 186(6_Supplement_2): S407 - S413. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. S. Berman, R. Hachamovitch, L. J. Shaw, J. D. Friedman, S. W. Hayes, L. E.J. Thomson, D. S. Fieno, G. Germano, P. Slomka, N. D. Wong, et al. Roles of Nuclear Cardiology, Cardiac Computed Tomography, and Cardiac Magnetic Resonance: Assessment of Patients with Suspected Coronary Artery Disease J. Nucl. Med., January 1, 2006; 47(1): 74 - 82. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. J. Dubinsky Coronary Artery Calcification Scoring Am. J. Roentgenol., December 1, 2005; 185(6): 1540 - 1541. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
