| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Cardiac Imaging |
1 From the Department of Radiology, FuWai Cardiovascular Institute and Hospital, Peking Union Medical College and Chinese Academy of Medical Sciences, Beijing, China (B.L., N.Z.); and the Department of Medicine, Division of Cardiology, Harbor-UCLA Medical Center and Saint John’s Cardiovascular Research Center, 1124 W Carson St, RB-2, Torrance, CA 90502 (S.S.M., J.C., S.C., H.B., M.J.B.). Received August 7, 2001; revision requested September 28; revision received November 26; accepted January 18, 2002. Address correspondence to M.J.B. (e-mail: mbudoff@rei.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Two hundred patients underwent electron-beam CT once and again 5 minutes later to evaluate interexamination variability of CAC score. Group 1 (104 patients) underwent scanning with use of an optimal EKG-triggering protocol (EKG triggering performed individually at the time of least coronary arterial motion during the cardiac cycle); group 2 (96 patients) underwent scanning with use of conventional 80% R-R interval triggering (the most common protocol with the electron-beam CT scanner). Interexamination, intraobserver, and interobserver variations of CAC measurements were compared between groups by using unpaired t tests for both Agatston and volumetric scores (in square millimeters).
RESULTS: Coronary arterial motion artifacts were found in 26% (27 of 104) versus 80% (77 of 96) of patients in groups 1 and 2, respectively (P < .0001). Intraobserver, interobserver, and interexamination variabilities in volumetric score were derived, with values of 1.2%, 9.2%, and 15.9% in group 1 and 2.1%, 11.3%, and 25.9% in group 2, respectively. Interexamination variabilities in both Agatston and volumetric score were significantly reduced with individualized EKG triggering, as compared with conventional triggering (P < .05), but intra- and interobserver variabilities were not (P > .05).
CONCLUSION: Optimal EKG triggering improves the reproducibility of CAC measurement by reducing coronary arterial motion artifacts.
© RSNA, 2002
Index terms: Computed tomography (CT), electron beam, 54.12118 • Coronary vessels, calcification, 54.81 • Coronary vessels, CT, 54.12118
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
For a diagnostic test to be useful, it must be demonstrated as accurate, clinically relevant, and reproducible. If electron-beam CT can demonstrate CAC sufficiently and reproducibly, then it can be useful for monitoring the progression and regression of coronary atherosclerosis by depicting changes in the quantity of atherosclerotic calcification. However, preliminary data show that poor interexamination reproducibility of CAC scoring was a major limitation in evaluating the progression of CAC at electron-beam (5–11) and dual-section helical CT (12).
Both intraobserver and interobserver variation in calcium scores at electron-beam CT have been shown to be extremely low, with reliability coefficients greater than 0.99 (13) and variabilities of 0%–10% (11,14–16). When two examinations are performed in the same patient in close temporal proximity, interexamination variation can be as high as 19%–49% (5–11,17) and is more pronounced in patients with low scores than in patients with high scores (9,18). Coronary arterial motion is one of the important causes of poor interexamination reproducibility; however, how big an effect, and, more important, how to reduce coronary arterial motion have not been well demonstrated (5–11,15–17).
To a certain degree, electrocardiographic (EKG) triggering ensures scanning of the entire heart, without gaps or overlap caused by transverse movement of the heart during a cardiac cycle. Some studies (19,20) have shown that optimal EKG triggering could reduce coronary arterial motion artifact. Optimal EKG triggering should be set at the end of systole (at the T wave at EKG), as compared with the 80% R-R interval (mid-diastole) during the cardiac cycle (20). We hypothesized that more reliable EKG triggering will decrease coronary arterial motion and probably yield further improvements in the reproducibility of CAC scoring. The aim of our study was to test our hypothesis that CT scanning during optimal EKG triggering could minimize image motion artifact and reduce interexamination CAC score variation.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Images from the first examination (also electron-beam CT) were visually evaluated for any CAC; if any was found, patients underwent a second examination for CAC with electron-beam CT. Only those patients with positive findings at the first examination (CAC in at least one of the three major coronary arteries) were eligible for this study. Other predefined exclusion criteria were excessive patient motion at one of the two examinations and poor cardiac EKG triggering, although no patients were excluded because of these problems. Because of the concern about variation in CAC quantification, every patient who underwent two consecutive cardiac CT examinations underwent the second examination after a 5-minute delay, without being moved between examinations, to minimize any positional differences.
Electron-Beam CT
Patients were examined with the use of a model C-150XL scanner (Imatron, South San Francisco, Calif). The procedure of CAC screening with electron-beam CT has been previously described (1–4). Thirty to 35 contiguous transverse images were obtained from the inferior margin of the right pulmonary arterial level (approximately 10 mm superior to the left main coronary artery [LM]) to the bottom of the heart. Each scan was obtained in a single breath hold by using 100-msec exposure and single-section-mode imaging. A section thickness of 3 mm, a field of view of 30 cm, and a matrix of 512 x 512 were used to reconstruct the raw image data, yielding a nominal pixel size of 0.59 x 0.59 mm2 and a voxel volume of 1.0 mm3. A sharp reconstruction filter was used to sensitively identify CAC.
Image acquisition was triggered to the 80% R-R interval (group 2) of EKG or to individually selected times based on the baseline heart rate (group 1) (Table 1). Our EKG-triggering protocol for group 1 was derived from our study of coronary arterial motion (20). We used the linear regression equation of Y = 0.3755 x X to obtain the least-motion point of the coronary artery during the cardiac cycle in patients with different heart rates, where Y represented the R-R percentage interval and X represented the baseline heart rate prior to scanning (20). This protocol, shown in Table 1, is optimal because the triggering time of EKG was set on the least-motion area (50 msec ahead of the least-motion point, to allow the 100-msec acquisition window to cover the area of least coronary arterial motion) and was determined individually, according to the patient’s baseline heart rate.
|
The Agatston score was calculated by the workstation (14). A lesion with a peak pixel intensity of 130–199 HU is assigned a coefficient value of 1, a peak pixel intensity of 200–299 HU is assigned a coefficient value of 2, a peak pixel intensity of 300–399 HU is assigned a coefficient value of 3, and a peak pixel intensity greater than or equal to 400 HU is assigned a coefficient value of 4. The total area (in square millimeters) of each calcified lesion is determined by using the software package and is multiplied by the value derived from the peak pixel intensity to yield a calcium score for that lesion. The sum of all calcium scores for each of the major epicardial arteries is determined as the total calcium score for the examination, for example, total Agatston score (in the patient) = 
(coefficient value · calcium area). We also used the calcium volume (in cubic millimeters) as the volumetric score.
Coronary Arterial Motion Artifact
Coronary arterial motion artifacts, caused by cardiac pulsation and appearing as long radiating streaks emitting from the vessel or its calcific plaques, impair accurate assessment of the cardiac images. Previous studies (20,21) have shown that motion velocity of the right coronary artery (RCA) was substantially faster than that of the left anterior descending (LAD) and left circumflex (LCX) coronary arteries and that more motion artifacts occurred in the RCA than in the LAD and LCX. To estimate the frequency of coronary arterial motion, we counted coronary arterial motion artifacts of the RCA in a section-by-section and patient-by-patient fashion. We also compared coronary arterial motion artifacts between groups 1 and 2. We sought an objective criterion that corresponded to visible motion of the coronary arteries on transverse electron-beam CT images. We defined coronary arterial motion as in-plane coronary arterial movement of a distance longer than its diameter.
Data Postprocessing and Statistics
The calcium score was determined on a vessel-by-vessel and patient-by-patient basis for both examinations by using the Agatston and volumetric scores (in square millimeters), respectively. For all patients, the percentage difference (variability) in calcium score between the two observers (interobserver), within observers (intraobserver), and between examinations (interexamination) were evaluated. Variability was defined as the absolute value of the difference in calcium scores divided by the mean calcium score and was expressed as a percentage; for example, interexamination variability (percentage) = [(examination 1 - examination 2)/mean calcium score (examination 1 + examination 2)] x 100. Variability was determined for both Agatston and volumetric scores. For the individual epicardial arteries (LM, LAD, LCX, and RCA), intraobserver, interobserver, and interexamination variability of calcium scores were calculated separately and combined in total scores. Interobserver, intraobserver, and interexamination agreements were measured by using the Pearson correlation coefficient. The absolute values of the difference in calcium scores between examinations were determined. Calcium score variability was also evaluated in different amounts in baseline score subgroups.
To determine the interexamination correlation in calcium scores, the data were log transformed (log10[mean calcium score]) to reduce skewness, and Bland-Altman plots were used to show relationships between variability and calcium distribution. The Wilcoxon matched-pairs test was used to evaluate the differences in examination results between the first and second examinations, by using the Agatston and volumetric score, respectively. Analyses of variance of Bonferroni t tests were used to compare variations in calcium scores among more than two vessels. Paired and unpaired Student t tests were used to identify differences in and between groups, respectively, whereas a P value of less than .05 indicated a significant difference. A 
2 analysis of contingency table tests was used to analyze the different rates of image motion artifact between groups 1 and 2.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
50) in group 1 was significantly lower than that in group 2 (P < .05), and interexamination variabilities in Agatston score subgroups of 51–400, 401–1,000 and greater than 1,000 were all lower in group 1 than in group 2; however, no significant differences were identified. In this study, we also evaluated motion artifacts, as shown in the RCA on transverse CT images (Fig 3). Data showed significantly fewer coronary arterial motion artifacts in patients with optimal EKG triggering in group 1 than in patients with conventional EKG triggering in group 2 (P < .0001) (Table 5).
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Early studies have demonstrated the mean variabilities of calcium scores, with electron-beam CT and use of the traditional 80% trigger for repeated scanning, as 29% (6), 37.2% (10), and 49% (9) and the limit of agreement as 0.91 to -0.94 (15). Hernigou et al (11) have shown intraobserver (n = 41), interobserver (n = 59), and interexamination (n = 50) variabilities as 2.8%, 3.1%, and 39.3%, respectively. All of these studies involved the conventional or defaulted EKG triggering protocol—for example, 80% R-R interval triggering with the electron-beam CT scanner, which corresponds to the protocol in group 2 (interexamination variability, 25.9%). With use of our optimal EKG triggering protocol, interexamination variability was reduced to 15.9% (P < .01).
With use of an 80% R-R trigger (group 2), interexamination variabilities in individual arterial scores were highest in the LM, followed by the RCA, LCX, and LAD (P = .28). Similar results were obtained in the study by Devries et al (9) and were probably due to use of the same EKG triggering method. In group 1, however, significantly lower interexamination variations were found in the LAD, LCX, and RCA, as compared with those in group 2 (P < .05 in all); this may have been due to less cardiac motion as the result of using a different EKG trigger. The finding that the largest score variation occurred in the LM may have been due in part to difficulty in delineating the exact junction between the LM, LAD, and LCX (16). For example, calcification near the junction of the LM bifurcation into the LAD and LCX may be assigned to the LM, whereas at the second reading or examination, the same lesion may be assigned to the LAD or LCX. Although this problem in assigning calcifications to a single artery should also increase interexamination variation for the LM, LAD, and LCX, interexamination variation in a patient’s total calcium scores would not be affected.
Poor reproducibility of CAC scoring at electron-beam CT is not principally the result of interobserver and intraobserver differences, as these variations are lower than interexamination differences. In the current study, excellent reproducibility of intra- and interobserver measurements was shown, with variation of median values as low as 0% and 2.4%, respectively. In other studies (5,13,18), intra- and interobserver reliability coefficients were very high (0.99 or more).
The reason for the good intraobserver reproducibility is that the region of interest of the automated measurement system in the workstation software package is applied to each coronary artery systematically from their ostial regions to distal portions on images. There are several sources of interobserver variability. First, some disagreements between observers exist in identifying the coronary arteries. Most commonly, in the LAD, some pericardial lesions or simple noise around the heart was considered CAC; in the LCX, mitral valve calcification or noise from the great cardiac vein or coronary sinus was misdiagnosed as CAC; in the RCA, distal segments coexist with the liver, and for this reason, image noise with attenuation values greater than 130 HU is easily identified in these areas, especially in patients who are obese. Image noise is sometimes incorrectly identified as CAC. Second, interobserver variability is caused by evaluation of ostial calcification. Calcification on the Valsalva sinus or aortic wall around the ostium of the LM or RCA were easily identified as ostial calcifications. These discrepancies in identifying the coronary arteries may be avoided with more careful measurement and more detailed information about coronary arterial anatomy.
There are some hypothetical reasons for explaining interexamination variations, such as cardiac position changes (15), coronary arterial motion artifacts (5–13), image noise (8), intraobserver and interobserver variations (13,15,16), and partial volume effects (6,8,12,13,15). However, the predominant effect is probably coronary arterial motion (23). Our data show that 80.2% of the patients in group 2 had coronary arterial motion artifacts (average, 10.6 sections per patient); however, only 26.0% of the patients in group 1 had coronary arterial motion artifacts (average, 1.0 section per patient) (P < .0001). We have previously demonstrated that the motion-free period is longer at end systole, as compared with end diastole (20,24). The reason for worse interexamination reproducibility of calcium scoring obtained with conventional EKG triggering is that more coronary arterial motion artifacts impair cardiac imaging.
Our data show that the variation in CAC score during repeated CT scanning is remarkably reduced with optimal EKG triggering, as compared with that obtained with conventional triggering at the 80% R-R interval. We found that interexamination variability in small calcific lesions (Agatston score 50) was significantly reduced by using optimal (group 1) rather than conventional (group 2) EKG triggering (P < .05). In one study (8) it was demonstrated that EKG triggering at a 40% R-R interval, as compared with conventional triggering (median variability, 8.9%), could reduce motion artifact and calcium score variability (median, 6.7%). Mao and colleagues (24) have demonstrated that 40% rather than 80% of R-R interval triggering was strongly recommended for electron-beam CT examinations for CAC. If EKG triggering was not used, either with helical scanning or with single-section scanning at helical CT, interexamination variability of calcium scoring increased (42.2%– 61.4%), as compared with EKG-triggered scanning (22.1%–25.4%) (25). Authors of another study (6) have indicated a method that could reduce interexamination variability—use of 6- rather than 3-mm-section scanning. This phenomenon is probably explained by increasing volume averaging. However, 6-mm-section scanning causes another problem—impaired sensitivity for detecting small calcific lesions—and has been shown to be substantially flawed (17,26).
New scoring methods, such as continuous volume scoring, have been introduced to decrease variations of CAC measurement (17). Our data show again that interexamination variation of calcium measurement was smaller with use of the volumetric score than with use of the Agatston score (P < .01). However, intra- and interobserver variabilities did not improve with use of the volumetric score (P > .05). In the study by Callister et al (17), the median percentage changes between two CAC quantification examinations were 13% for the volumetric score and 19% for the traditional score, and the overall reduction in error with the volumetric score was 40%. This phenomenon can be explained by the hypothesis that the peak pixel intensity in calculating the Agatston score varied between examinations (8,17).
Other effects of examination reproducibility, such as image noise, have been discussed (8,25). Image noise at CT is related to low radiation dose, which might hinder the detection of small calcified plaques (27). Achenbach et al (8) showed that Agatston score variability in the highest tercile of image noise (>28.5 HU) was 20.0% ± 33.3 (median, 11.7%), which was significantly higher than the lowest tercile of image noise (<25.5 HU) in 13.5% ± 30.9 (median, 4.6%). It is impossible to change the tube current or voltage to decrease image noise with the use of currently available electron-beam CT scanners. In patients who are obese or have small amounts of CAC, image noise can impair examination reproducibility.
For images free of cardiac motion artifacts, acquisition times shorter than 19.2 msec are necessary (28). With optimal EKG triggering it is possible to obtain cardiac images in the least-motion area in a cardiac cycle; and the minimum scanning speed or acquisition time needed to reduce motion artifacts can be decreased to 50–70 msec for a patient baseline heart rate of 50–100 bpm (20). Because of the longer exposure time of single- or multisection helical CT, an interval of 50% was chosen to ensure image acquisition just before substantial motion artifacts arose from systolic cardiac movement (25). This still does not limit the poor reproducibility of helical CT due to long exposure times, as compared with that of electron-beam CT (25).
Because the heart rate changes during scanning in respiratory suspension, EKG misregistration will cause inconsistencies in section acquisition and will result in gaps or overlaps between sections. The influence of a changing heart rate and arrhythmias on CAC assessment need to be evaluated. Scan width, field of view, and variation in patient circumference and position in the field of view can also affect measurement (18,29).
Detection of small amounts of calcium is limited with currently available electron-beam CT scanners. Lesions 2 and 3 pixels in size were reproduced at both examinations in only 21% of cases, and the percentage difference in their scores averages 90.0% (10), which corresponds to scores in group 2 in the current study (100.6% average percentage difference). Although the optimal EKG protocol was used in the current study, interexamination variability of calcium score for quantifying small lesions was still as high as 38.5%. Changing collimation by using 1.5-mm-section scanning instead of conventional 3-mm-section scanning for image acquisition may be necessary for detecting small amounts of calcium. Multi–detector row CT can produce four 1-mm collimated sections with use of 3.6-mm/sec table speed and 500-msec rotation (30). Retrospective EKG gating of continuous helical CT acquisitions allows image reconstruction at any time during the cardiac cycle. While these theoretical advantages of multi–detector row CT may be helpful to detect small calcified lesions, a relatively slow acquisition time with multi–detector row helical CT (250 msec per section) versus that with electron-beam CT (100 msec per section) precludes it being a feasible alternative to electron-beam CT for CAC measurement (30,31), especially in patients with heart rates faster than 70 bpm. Authors of a recent reproducibility study of CAC in which helical CT was used (23) have demonstrated interexamination variability of 40% ± 58. Also, because only a fraction of the acquired data is used, this increases the radiation dose (22). Multiple analyses at different points in the cardiac cycle will increase image processing time and interreader variability.
Interexamination variation remains an important limitation of electron-beam CT in the examination of asymptomatic patients. Interexamination variations due to positional differences in patients between first and subsequent examinations is unavoidable but to our knowledge have not yet been shown to be problematic for tracking the progression or regression of CAC over time with electron-beam CT (32). Optimal EKG-triggered data acquisition at electron-beam CT plays an important role in the reduction of study variations. Use of our trigger reduces interexamination variability by more than 33% (from 29.2% to 20.1%), and volumetric scoring further reduces variability by 21% (from 20.1% to 15.9%).
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, B.L., M.J.B.; study concepts, B.L., N.Z., S.S.M., J.C., S.C., H.B., M.J.B.; study design, B.L., N.Z., S.S.M., J.C., S.C., H.B.; literature research, B.L., N.Z.; clinical studies, B.L., J.C., S.C., H.B., M.J.B.; data acquisition, B.L., N.Z., J.C., S.C., H.B.; data analysis/interpretation, B.L., N.Z.; statistical analysis, B.L., N.Z.; manuscript preparation, B.L., N.Z.; manuscript definition of intellectual content, B.L., M.J.B.; manuscript editing, B.L., N.Z., M.J.B.; manuscript revision/review, B.L., N.Z., S.S.M., M.J.B.; manuscript final version approval, B.L., M.J.B.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  I. Gottlieb and J. A.C. Lima Screening High-Risk Patients With Computed Tomography Angiography Circulation, March 11, 2008; 117(10): 1318 - 1332. [Full Text] [PDF]
|
|
|
|
 |
|
 A. Rutten, I. Isgum, and M. Prokop Coronary Calcification: Effect of Small Variation of Scan Starting Position on Agatston, Volume, and Mass Scores Radiology, December 1, 2007; 246(1): 90 - 98. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Budoff, S. Achenbach, R. S. Blumenthal, J. J. Carr, J. G. Goldin, P. Greenland, A. D. Guerci, J. A.C. Lima, D. J. Rader, G. D. Rubin, et al. Assessment of Coronary Artery Disease by Cardiac Computed Tomography: A Scientific Statement From the American Heart Association Committee on Cardiovascular Imaging and Intervention, Council on Cardiovascular Radiology and Intervention, and Committee on Cardiac Imaging, Council on Clinical Cardiology Circulation, October 17, 2006; 114(16): 1761 - 1791. [Full Text] [PDF]
|
|
|
|
 |
|
 D. T. Boll, A. S. Bossert, A. J. Aschoff, M. H. Hoffmann, and R. C. Gilkeson Synergy of MDCT and Cine MRI for the Evaluation of Cardiac Motility Am. J. Roentgenol., June 1, 2006; 186(6_Supplement_2): S379 - S386. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. R. Nandalur, E. Baskurt, K. D. Hagspiel, M. Finch, C. D. Phillips, S. R. Bollampally, and C. M. Kramer Carotid Artery Calcification on CT May Independently Predict Stroke Risk Am. J. Roentgenol., February 1, 2006; 186(2): 547 - 552. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Clouse, J. Chen, H. M. Krumholz, M. E. Clouse, J. Chen, and H. M. Krumholz Noninvasive Screening for Coronary Artery Disease With Computed Tomography Is Useful Circulation, January 3, 2006; 113(1): 125 - 146. [Full Text] [PDF]
|
|
|
|
|
|
A. L. Daniell, N. D. Wong, J. D. Friedman, N. Ben-Yosef, R. Miranda-Peats, S. W. Hayes, X. Kang, M. G. Sciammarella, L. de Yang, G. Germano, et al. Concordance of Coronary Artery Calcium Estimates Between MDCT and Electron Beam Tomography Am. J. Roentgenol., December 1, 2005; 185(6): 1542 - 1545. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. B. Sevrukov, J. M. Bland, and G. T. Kondos Serial Electron Beam CT Measurements of Coronary Artery Calcium: Has Your Patient's Calcium Score Actually Changed? Am. J. Roentgenol., December 1, 2005; 185(6): 1546 - 1553. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Katoh, J. E. Wildberger, R. W. Gunther, and A. Buecker Malignant Right Coronary Artery Anomaly Simulated by Motion Artifacts on MDCT Am. J. Roentgenol., October 1, 2005; 185(4): 1007 - 1010. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Schlosser, P. Hunold, A. Schmermund, H. Kuhl, K.-U. Waltering, J. F. Debatin, and J. Barkhausen Coronary Artery Calcium Score: Influence of Reconstruction Interval at 16-Detector Row CT with Retrospective Electrocardiographic Gating Radiology, November 1, 2004; 233(2): 586 - 589. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Lu, N. Zhuang, S.-S. Mao, J. Child, S. Carson, and M. J. Budoff Baseline Heart Rate-adjusted Electrocardiographic Triggering for Coronary Artery Electron-Beam CT Angiography Radiology, November 1, 2004; 233(2): 590 - 595. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. J. Schoepf, C. R. Becker, B. M. Ohnesorge, and E. K. Yucel CT of Coronary Artery Disease Radiology, July 1, 2004; 232(1): 18 - 37. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. E. Hokanson, T. MacKenzie, G. Kinney, J. K. Snell-Bergeon, D. Dabelea, J. Ehrlich, R. H. Eckel, and M. Rewers Evaluating Changes in Coronary Artery Calcium: An Analytic Method That Accounts for Interscan Variability Am. J. Roentgenol., May 1, 2004; 182(5): 1327 - 1332. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Desjardins and E. A. Kazerooni ECG-Gated Cardiac CT Am. J. Roentgenol., April 1, 2004; 182(4): 993 - 1010. [Full Text] [PDF]
|
|
|
|
|
|
W. Stanford, B. H. Thompson, T. L. Burns, S. D. Heery, and M. C. Burr Coronary Artery Calcium Quantification at Multi-Detector Row Helical CT versus Electron-Beam CT Radiology, February 1, 2004; 230(2): 397 - 402. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Vliegenthart, B. Song, A. Hofman, J. C. M. Witteman, and M. Oudkerk Coronary Calcification at Electron-Beam CT: Effect of Section Thickness on Calcium Scoring in Vitro and in Vivo Radiology, November 1, 2003; 229(2): 520 - 525. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||
