| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, Hiroshima University School of Medicine, 1-2-3,
Kasumi-cho, Minami-ku, Hiroshima, 734-8551, Japan
2
Department of Radiology, Mazda Hospital, 2-15, Aosakiminami, Huchu-cho,
Aki-gun, Hiroshima, 735-0017, Japan.
Received April 18, 2001;
accepted after revision June 11, 2001.
Address correspondence to J. Horiguchi.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. First, 20 volunteers were scanned using multidetector CT. We compared the signal-to-noise ratio in the heart, motion artifacts at the heart border, and the highest CT values in the regions of the coronary arteries using single-sector and multisector reconstruction algorithms. Next, 60 patients with coronary calcified deposits underwent both multidetector CT and electron beam CT. We compared coronary calcium scores determined with multidetector CT using the two algorithms (thresholds of 90 and 130 H) with those determined using electron beam CT.
RESULTS. The signal-to-noise ratio was higher and motion artifacts were reduced when we used the multisector algorithm. The highest CT value in the region of the coronary arteries exceeded 90 H in one of 55 arteries on the multisector algorithm images and 17 of 55 arteries on single-sector algorithm images (chi-square test, p < 0.01). In coronary calcium scoring, correlation coefficients ranged from 0.920 to 0.992 (Pearson's product moment) and from 0.932 to 0.969 (Spearman's rank correlation coefficient).
CONCLUSION. Multidetector CT with a retrospective ECG-gating algorithm (multisector reconstruction) produced cardiac images with fewer motion artifacts and showed a high correlation with coronary calcium scores determined using electron beam CT. Therefore, multidetector CT is a potential tool for coronary calcium scoring.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We used the following methodology for multisector reconstruction: First, ECG-gated data were collected during multiple heartbeats by helical or cine CT using a multirow detector. Second, ECG-gating data sets corresponding to single-sector reconstructions were divided into several sectors (n, 2-4). We then selected sectors of the same phase window on the basis of the ECG information and combined the sectors for an image reconstruction (Fig. 1). The temporal resolution achieved with the combination of the gantry rotation speeds of 0.8 and 1.0 sec ranged from 133 to 300 msec (Fig. 2).
|
|
The sample size consisted of 20 male volunteers who were between 26 and 55 years old (mean age ± SD, 39 ± 6 years). Informed consent was obtained from all participants. For multidetector CT scans, we obtained volumetric data for the heart in the helical mode with scan parameters of a 2.5-mm collimation width x 4 detectors, a helical pitch of 1.0, 120 kV, and 200 mA. Although higher temporal resolution can be achieved by combining two gantry speeds, we chose a speed of 0.8 sec per rotation regardless of the participant's heart rate for two reasons. First, using a pitch of 1.0 may make it difficult to cover the entire heart with a single breath-hold at the rotation speed of 1.0 sec per rotation. Second, we found it difficult to accommodate a participant's heart rate to an ideal rotation speed because the heart rate could easily vary during the approximately 40-sec breath-hold. Diastolic phase images were reconstructed from the same raw data sets for both single-sector and multisector reconstructions using an external workstation equipped with prototypical software designed specifically for cardiac image reconstruction. Reconstruction was performed with a 512 x 512 matrix, a 26-cm field of view, and a pixel area of 0.26 mm2. To obtain motionless images, we set the center of capture time to 80% of the R-R wave interval on single-sector reconstructions and the end of the temporal window to 80% of the R-R wave interval on multisector reconstructions. These images then were transferred to another external workstation (Advantage Windows 3.1, General Electric Medical Systems) for the analysis.
Regions of interest of approximately 100 mm2 each were set to the areas that were almost free of motion artifacts at three levels—the ascending aorta at the level of the left main coronary trunk, left atrium at the mid ventricular level, and left ventricle at the base of the heart. We measured the means and the standard deviations (SDs) of CT values in the regions of interest and then calculated S/N, which we defined as the value mean divided by the SD. These were compared with the two reconstructions using Student's t test. We chose two representative levels of the heart, the right mid coronary artery and the left ventricular apex. The peak CT values at the heart border attributable to motion artifacts were measured and compared using Student's t test.
Motion artifacts in the region of the coronary arteries were evaluated. Of the 60 coronary arteries in our 20 volunteers, five arteries were excluded from the study because of the presence of calcified deposits. Using a workstation, we extracted pixels encompassing each coronary artery and pericoronary fat as far from the origin as possible and then gathered the pixels into a voxel (Fig. 3). We performed this procedure on the remaining 55 coronary arteries (right coronary, left main to left anterior descending, and left circumflex) under study and measured peak CT value in these voxels. The coronary arteries were categorized into two groups, one group with peak CT values of less than 90 H and the other with peak CT values of more than or equal to 90 H. We compared the peak CT values in the coronary arteries of the two reconstructions using the chi-square test.
|
We then sought to compare the total calcium scores derived from electron beam CT with that derived from multidetector CT and to verify factors affecting the correlation. For this part of the study, 60 patients (45 men and 15 women) who were between 49 and 80 years old (mean age, 69 ± 9 years) with coronary calcium deposits that had been confirmed by electron beam CT were included. Eleven patients had a history of ischemic heart disease, and eight had diabetes mellitus. Informed consent was obtained from all patients.
We determined the calcium score using the Agatston score [1] for electron beam CT and multidetector CT and using a modified Agatston score [5] for multidetector CT. Using the Agatston method, we defined the regions of interest by vessel and slice with the threshold option for pixels greater than 130 H so that we could measure the area and peak density of plaques. An area of at least 0.52 mm2 (2 pixels) was multiplied by one of the following factors: a factor of 1 for peak plaque densities of 130-199 H, a factor of 2 for densities of 200-299 H, a factor of 3 for densities of 300-399 H, and a factor of 4 for densities of 400 H or greater. The total calcium score was calculated as the sum of the individual lesion scores in all coronary arteries. The modified Agatston score was obtained using a CT density threshold for calcium of 90 H instead of 130 H. The multidetector CT scan protocol was the same as described earlier.
An electron beam CT scanner (C-150 XL; Imatron, South San Francisco, CA) was used. Imaging parameters included a scan of 100 msec, 35-40 continuous gapless slices of 3-mm thickness, 130 kV, and 625 mA. The single-section images were obtained by ECG triggered to 80% of the R-R wave interval. The matrix size (512 x 512) and the field of view (26 cm) were the same as those used for multidetector CT.
Consecutive diastolic phase images of electron beam CT and multidetector CT (single-sector and multisector reconstructions) were transferred to a computer (Accuview; Imatron) for quantification of coronary artery calcification. Using software for quantification of coronary artery calcification (Accuscore; Imatron), we calculated the Agatston score of electron beam CT and the Agatston and modified Agatston scores of multidetector CT individually.
Because the reconstruction increments of multidetector CT (both single-sector and multisector) varied according to the patient's heart rate, we applied corrections to the calculated Agatston and modified Agatston scores of multidetector CT as shown in the following formula: corrected score = calculated score x mean slice interval (mm) / 3 (mm).
We compared total calcium scores produced with electron beam CT and corrected scores produced with multidetector CT using both Pearson's product moment and Spearman's rank correlation coefficients.
We propose that the values calculated using the following formula reflect an approximation of the coronary artery calcification score between multidetector CT and electron beam CT: approximate value = absolute (multidetector CT — electron beam CT) / electron beam CT.
Using Student's t test, the values were compared with the scores of the single-sector reconstruction with a threshold of 90 H (single-sector 90), single-sector reconstruction with a threshold of 130 H (single-sector 130), multisector reconstruction with a threshold of 90 H (multisector 90), and multisector reconstruction with a threshold of 130 H (multisector 130).
Finally, targeting to multisector 130, we categorized the 60 patients into two groups of 30 each by the median number of the heart rate, temporal resolution, the value temporal resolution divided by R-R wave interval, and Agatston score of electron beam CT. We then used Student's t test to compare the approximation values between the two groups separated by each factor.
Using a 10-cm pencil-shaped ionization chamber in a 32-cm-diameter cylindrical phantom, we measured the CT dose index at the center and at four locations in the periphery of the phantom (Radiation Monitor Controller 9015; Radocal, Monrovia, CA). The weighted CT dose index was determined to be two thirds of the mean peripheral CT dose index and one third of the central CT dose index. Dose-length-product was calculated as weighted CT dose index x collimation x number of rotations.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
On measuring the CT value of motion artifacts at the heart border, we found the peak CT values of pixels at the sections of the mid right coronary artery and left ventricular apex were 165 ± 50.6 H and 140.0 ± 44.3 H on single-sector reconstructions and 111.0 ± 23.5 H and 113.0 ± 30.4 H on multisector reconstructions. We found a statistically significant difference between the peak CT values of the two reconstructions (p < 0.01). The difference between the two in the number of coronary arteries with peak CT values above 90 H was also statistically significant (p < 0.01; single-sector, 18/55 or 32.7%; multisector, 1/55 or 1.8%).
In the 60 patients with coronary artery calcium, Pearson's product moments between the scores derived with electron beam CT and the corrected scores derived with multidetector CT were 0.920 in single-sector 90, 0.977 in single-sector 130, 0.992 in multisector 90, and 0.988 in multisector 130. Spearman's rank correlation coefficients between the scores were 0.932 in single-sector 90, 0.935 in single-sector 130, 0.969 in multisector 90, and 0.957 in multisector 130.
The equations of linear correlation between electron beam CT and multidetector CT were as follows (Figs. 4,5,6,7): single-sector 90 score = 29.8 + 1.26 x electron beam CT score (r = 0.930); single-sector 130 score = 26.8 + 0.901 x electron beam CT score (r = 0.977); multisector 90 score = 8.0 + 1.12 x electron beam CT score (r = 0.992); and multisector 130 score = 4.0 + 0.849 x electron beam CT score (r = 0.988).
|
|
|
|
The approximate values were 0.53 ± 0.70 in single-sector 90, 0.32 ± 0.41 in single-sector 130, 0.27 ± 0.30 in multisector 90, and 0.25 ± 0.19 in multisector 130. There were statistically significant differences in the following values: The values for single-sector 90 were higher than those for single-sector 130, and the values for single-sector 130 were higher than those of multisector 90 (p = 0.05). However, no statistical difference was found between the values for multisector 90 and multisector 130.
In multisector 130, the median values of the heart rate, temporal resolution, temporal resolution divided by R-R wave interval, and Agatston score of electron beam CT were 70 beats per minute, 271 msec, 0.30, and 264, respectively (Table 1). These approximate values were not affected by the first three factors (p > 0.05), although these values tended to be low in the lower heart rate patient group. The values in the group of patients with lower calcium scores were significantly higher than those in the group with higher calcium scores (p < 0.05). Pearson's product moments between the electron beam CT score and multisector 130 score were 0.844 in the lower calcium group and 0.985 in the higher calcium group, and Spearman's rank correlation coefficients of electron beam CT score and multisector 130 were 0.874 in the lower calcium group and 0.899 in the higher calcium group.
|
For CT of the heart with z-axis coverage of 10.5 cm, the weighted CT dose index was 3.67 mGy and dose-length-product was 38.53 mGy cm with electron beam CT. The weighted CT dose index was 49.42 mGy and dose-length-product was 496.67 mGy cm with multidetector CT. The multidetector CT in our protocol led to an effective dose that was 13 times higher than that of electron beam CT.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
On electron beam CT scans, hyperattenuating noise is sometimes difficult to distinguish from calcification. Although it is impossible to change the scanning parameters of electron beam CT, we can increase the tube current of single-detector or multidetector helical CT and subsequently reduce image noise [10]. This option may be helpful in detecting coronary artery calcification in obese patients or in patients with small amounts of coronary artery calcification (Fig. 8A,8B).
|
|
Assessments of coronary artery calcification using nonhelical conventional CT have been reported [11, 12]. One study on helical CT has shown that, even with accelerated scan times, calcified deposits are often blurred because of cardiac motion, and small amounts of calcium may not be seen [13]. However, recent studies have shown good results using double-helix helical CT [14], ECG-triggered single-slice CT [10], and retrospective-gated helical CT [15].
Multidetector CT allows acquisition of thinsection volumetric data in a single breath-hold, and multidetector CT equipped with an ECG provides two options for data acquisition method—prospective ECG triggering (step and shoot) and retrospective ECG gating. The prospective triggering technique with multidetector CT using segmented data of a gantry rotation provides a temporal resolution of 250-533 msec with single-sector reconstruction (depending on gantry rotation speed and reconstruction algorithm). This technique results in less radiation exposure than retrospective reconstruction but is sensitive to changes in a patient's heart rate during the examination.
Use of retrospective gating exposes the patient to excessive radiation caused by oversampling. Retrospective gating requires continuous scanning through the entire heart, although only part of the sampling data are used for reconstruction of the CT images. However, retrospective reconstruction has some advantages. A recent study of multidetector CT found that, with slice-by-slice acquisition, it remains difficult to reproduce minute calcified plaque with electron beam or with multidetector CT [16]. In contrast, phantom results using retrospective reconstruction [17,18,19] have suggested that reproducibility, which is an important factor for the quantification of coronary artery calcification, could be improved by overlapping slice reconstruction [20]. In addition, retrospective reconstruction is less sensitive to heart rate changes during the examination. Moreover, a recent study has shown that in-plane coronary arterial motion velocity measured by electron beam CT varies considerably during the cardiac cycle, and variations among individual patients were high [21]. Unlike the data acquisition process used with a fixed triggering parameter, the process allows us to select and reconstruct images with fewer motion artifacts at any desired cardiac cycle by retrospective reconstruction. The parts of raw data used for reconstruction of diastolic images differ according to the capture time and data acquisition (prospective triggering or retrospective gating) methods. In electron beam CT, prospective ECG-triggering at 80% at the R-R wave interval has been used [1]. In helical CT techniques, prospective triggering at 50% [10], image selection (from approximately 10 frames per cardiac cycle) using ECG data and visual optimization [15], and prospective triggering at 450 msec before the next R wave [16] have been reported. In our protocol, we designed the reconstruction of motionless images so that the center of capture time corresponded to 80% of the R-R wave interval on single-sector reconstructions and the end of the temporal window was set to 80% of the R-R wave interval on multisector reconstructions. However, determining the optimal gating times on multisector reconstructions under inconstant temporal resolution and variable heart rates still remains difficult.
Partial-volume effect can be reduced when the slice thickness becomes thinner. In this sense, the thin-slice images may provide a more accurate coronary artery calcification scoring. Our protocol was designed to scan at the pitch of 1.0 with a collimation width of 2.5 mm. A larger pitch provides images of 3-mm slice thickness and shortens the breath-hold time at the expense of image quality. Although we do not know which is best suited for an accurate quantification of coronary artery calcification, a larger pitch would be an alternative.
Since the study in which Broderick et al. [5] applied a new threshold for calcium at 90 H with a helical CT and found a good association with coronary artery disease, 90 H has been used for the threshold of calcium in helical CT examinations [10, 14,15,16]. However, to our knowledge, no studies investigating the CT value of motion artifacts on helical CT have been reported. Our study has shown that retrospective ECG-gating multidetector CT provides images with high S/N and fewer motion artifacts. Artifacts in the region of the coronary arteries in multisector reconstructions rarely exceed a CT value of 90 H, which suggests that the use of the threshold of 90 H for the detection and quantification of calcium is an option.
Our correlation study showed good agreement for coronary artery calcification scores between electron beam CT and multidetector CT, especially using multisector reconstruction with thresholds of both 90 H and 130 H. However, according to the results of the calculation of approximation values, the scores produced by multidetector CT with a threshold of 130 H were more similar to those on electron beam CT than those scores produced by multidetector CT with 90 H.
Although the reasons for this finding are not clear, there are two possibilities to be considered. The first is that multidetector CT with a threshold of 90 H could reveal subtle calcification that was missed with electron beam CT. One group of researchers has shown, using an anthropomorphic phantom, that a 1-mm calcification not distinguishable from noise with electron beam CT was depicted with multidetector CT for the heart at rest and for heart rates ranging from 50 to 120 beats per minute (Ulzheimer S et al., presented at the Radiological Society of North America meeting, Chicago, November 2000). This study may suggest that, in evaluating coronary artery calcification by multidetector CT, the choice of a threshold for calcium correlating best to electron beam CT is not necessarily the ideal method. The second possibility is that the coronary artery calcification scores produced by multidetector CT tended to be high because artifacts arising from coronary artery calcification on multidetector CT scans were more noticeable than those on electron beam CT scans (Fig. 9A,9B). We believe that future improvement of temporal resolution of multidetector CT may further justify the use of the thresholds of 90 H. Further clinical examination is desirable to verify whether lesions with subtle calcified deposits delineated by multidetector CT really can be regarded as a sign of soft-tissue plaque in coronary arteries.
|
|
No significant differences were seen between the groups of patients with high and low coronary artery calcification scores in factors such as heart rate, temporal resolution, and the value temporal resolution divided by R-R wave interval in approximating coronary artery calcification scores found with multidetector CT and those found with electron beam CT. Approximation of the scores was higher in the group of patients with high coronary artery calcification scores. Similar results for helical CT have been reported by another group of researchers who found that the percentage of absolute difference determined as 100%*Absolute (helical CT — electron beam CT) / (mean of helical CT and electron beam CT) was much higher for individuals with low scores than it was for those with high scores (Stanford W et al., presented at the Radiological Society of North America meeting, Chicago, November 2000).
We were able to obtain satisfactory results in quantification of coronary artery calcification with multisector reconstruction, but some problem areas need to be addressed.
Temporal Resolution
Theoretically, temporal resolution of 133-300 msec can be achieved using
multisector reconstruction by combining two gantry rotation speeds. In
reality, the temporal resolution in our study ranged from 133 to 533 msec
because we used only one gantry rotation speed, 0.8 sec per rotation. For
images free of motion artifacts, acquisitions of less than 41.8 msec are
necessary [22]. However, such
a short capture time is impossible even with electron beam CT. To obtain
images with reduced motion artifacts, the relatively slow myocardial motion of
diastolic phase is chosen. Becker et al.
[8] indicated that motion-free
images of the heart could be obtained with 500, 250, and 100 msec scanning
times in patients with heart rates slower than 50, 70, and 110 beats per
minute, respectively. Faster gantry rotation speed is desirable for
improvement of temporal resolution as well as for a shorter breath-hold
time.
Reconstruction Increment
In single-sector helical CT, retrospective ECG gating has reduced motion
artifacts by improving temporal resolution
[23], and the new
interpolation algorithm allows image reconstruction at any z-position
and at any phase of the cardiac cycle
[24]. However, in multisector
reconstruction, increments are variable according to the heart rate, resulting
in slice gaps and overlaps. Reconstruction increments can be easily changed
during the examination with breathholds of approximately 30-40 sec at deep
inspiration [25]. We applied
corrections of calculated Agatston scores in consideration of mean slice
intervals on single-sector and multisector reconstructions. A new
z-interpolation technique allowing equidistant diastolic
reconstruction at any table position is needed
[26].
Radiation Exposure
Low spiral pitch scanning has the advantages of high spatial resolution and
reproducibility. However, it results in increased radiation exposure.
ECG-modulated dose exposure (tube current) was attempted, which resulted in a
50% dose reduction without compromising image quality for diastolic
reconstructions (Ohnesorge BM et al., presented at the Radiological Society of
North America meeting, Chicago, November 2000). For this technique to be
established as a screening tool for coronary artery calcification, this issue
must be considered.
In summary, multidetector CT using a retrospective ECG-gating reconstruction algorithm permits acquisition of cardiac images with high S/N ratios and fewer motion artifacts. The density of coronary arteries without calcified deposits rarely exceeds the threshold of 90 H. In the part of our study with patients, multidetector CT findings showed good correlation with electron beam CT findings for coronary artery calcification scoring. We conclude that multidetector CT holds the promise of having an important impact on coronary artery calcification assessment.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  J. Horiguchi, H. Yamamoto, N. Hirai, Y. Akiyama, C. Fujioka, K. Marukawa, H. Fukuda, and K. Ito Variability of repeated coronary artery calcium measurements on low-dose ECG-gated 16-MDCT. Am. J. Roentgenol., July 1, 2006; 187(1): W1 - W6. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Horiguchi, Y. Shen, Y. Akiyama, N. Hirai, K. Sasaki, M. Ishifuro, T. Nakanishi, and K. Ito Electron Beam CT Versus 16-MDCT on the Variability of Repeated Coronary Artery Calcium Measurements in a Variable Heart Rate Phantom Am. J. Roentgenol., October 1, 2005; 185(4): 995 - 1000. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Horiguchi, H. Yamamoto, Y. Akiyama, N. Hirai, K. Marukawa, H. Fukuda, and K. Ito Variability of Repeated Coronary Artery Calcium Measurements by 16-MDCT with Retrospective Reconstruction Am. J. Roentgenol., June 1, 2005; 184(6): 1917 - 1923. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. S. Halliburton, A. E. Stillman, M. Lieber, J. M. Kasper, S. A. Kuzmiak, and R. D. White Potential Clinical Impact of Variability in the Measurement of Coronary Artery Calcification with Sequential MDCT Am. J. Roentgenol., February 1, 2005; 184(2): 643 - 648. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Yamamuro, E. Tadamura, S. Kubo, H. Toyoda, T. Nishina, M. Ohba, R. Hosokawa, T. Kimura, N. Tamaki, M. Komeda, et al. Cardiac Functional Analysis with Multi-Detector Row CT and Segmental Reconstruction Algorithm: Comparison with Echocardiography, SPECT, and MR Imaging Radiology, February 1, 2005; 234(2): 381 - 390. [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]
|
|
|
|
|
|
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. Horiguchi, H. Yamamoto, Y. Akiyama, K. Marukawa, N. Hirai, and K. Ito Coronary Artery Calcium Scoring Using 16-MDCT and a Retrospective ECG-Gating Reconstruction Algorithm Am. J. Roentgenol., July 1, 2004; 183(1): 103 - 108. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. H. Mahnken, J. E. Wildberger, J. Simon, R. Koos, T. G. Flohr, S. Schaller, and R. W. Gunther Detection of Coronary Calcifications: Feasibility of Dose Reduction with a Body Weight-Adapted Examination Protocol Am. J. Roentgenol., August 1, 2003; 181(2): 533 - 538. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Saito, M. Saito, S. Komatu, and K. Ohtomo Real-Time Four-dimensional Imaging of the Heart with Multi-Detector Row CT RadioGraphics, January 1, 2002; (2002) 8. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK |
