AJR AJR-based Continuing Ed for Technologists
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Fei, X.
Right arrow Articles by Li, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Fei, X.
Right arrow Articles by Li, K.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
Hotlight (NEW!)
Right arrow
What's Hotlight?
DOI:10.2214/AJR.07.2653
AJR 2008; 191:43-49
© American Roentgen Ray Society

Original Research

64-MDCT Coronary Angiography: Phantom Study of Effects of Vascular Attenuation on Detection of Coronary Stenosis

Xiaolu Fei1, Xiangying Du1, Qi Yang1, Yun Shen2, Pengyu Li1, Jingmin Liao2 and Kuncheng Li1

1 Department of Radiology, Xuanwu Hospital of Capital Medical University, 45 Changchun St., Xuanwu District, Beijing 100053, China.
2 CT Laboratory of GE Healthcare, Beijing Economic and Technology Development Area, Beijing, China.

Received May 31, 2007; accepted after revision January 20, 2008.

 
Supported by grant Z0005190042691 from Beijing Municipal Science and Technology Commission.

Address correspondence to K. Li (kuncheng.li{at}gmail.com).


Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
OBJECTIVE. The purpose of this study was to investigate the effects of vascular attenuation on the accuracy of stenosis evaluation with 64-MDCT coronary angiography.

MATERIALS AND METHODS. A pulsating cardiac phantom was used to simulate the beating heart and coronary arteries of 5 and 3 mm in diameter with three degrees of stenosis (25%, 50%, and 75%) at a heart rate of 55 beats per minute. Coronary vascular enhancement had four attenuation levels: low, 200 H; moderately low, 300 H; moderately high, 350 H; and high, 500 H. Cardiac scans were obtained with 64-MDCT. Percentage stenosis, plaque area, and plaque density were measured on axial images.

RESULTS. For 50% and 75% stenosis in 5-mm vessels, there were no significant differences among the four attenuation groups. For 50% and 75% stenosis in 3-mm vessels, significant underestimation of percentage stenosis occurred in the high-attenuation group compared with the moderate- and low-attenuation groups (p < 0.05). For 25% stenosis in 5-mm vessels, low attenuation led to significant overestimation of degree of stenosis compared with the moderate and high attenuation levels (p < 0.05). None of the instances of 25% stenosis in 3-mm vessels were detected in the high-attenuation group. Underestimation was found only for 3-mm vessels. For 75% stenosis, all plaques were detected irrespective of contrast attenuation and vessel size.

CONCLUSION. Use of higher attenuation leads to a significant underestimation of stenosis in smaller vessels. Lower attenuation leads to slight and clinically acceptable overestimation of stenosis. The optimal vascular attenuation for stenosis detection in coronary 64-MDCT angiography is approximately 350 H.

Keywords: cardiac imaging • contrast medium • coronary CT angiography • stenosis detection • vascular attenuation


Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
MDCT has been used increasingly for coronary imaging in the noninvasive assessment of coronary artery stenosis [1–3]. At MDCT, including 64-MDCT, slice thickness can be 0.5 mm, and temporal resolution can reach 44 milliseconds with multisector reconstruction, further improving imaging of coronary stenosis at CT angiography (CTA) [1]. Contrast-enhanced coronary CTA with MDCT has become a strong candidate for a noninvasive imaging-based tool for assessment of cardiac disease. Accordingly, the proper use of iodinated contrast media is an important factor for ensuring accurate imaging results at coronary CTA.

The optimal parameters for a contrast medium, such as total iodine dose, iodine concentration, and injection rate and duration, have been extensively studied [4–8]. The effects of vascular attenuation at coronary CTA on the accuracy of coronary stenosis detection also have been studied [9, 10], but no consistent conclusion has been made, to our knowledge. The aim of this study was to use a pulsating cardiac phantom to assess the effects of different attenuation levels on the accuracy of stenosis evaluation at coronary CTA.


Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Phantom
In this study, we used a pulsating cardiac phantom developed in a CT laboratory (Fig. 1A, 1B). The design of the cardiac phantom was based on the following four characteristics [11, 12]. First, the pulsating phantom was made of two chambers for ECG gating information. The phantom contained an interior trigger and an exterior trigger. Second, the driver operated with a servomotor for changing the shape of the simulative left ventricle in 3D (x, y, and z directions) with 16 preset heart types. The fast pumping, fast filling, and slow filling phases of a real heart in a cardiac cycle were incorporated into the driver sequence, including the shift of a patient's heartbeat or irregular pulse (heartbeat, 0–120 beats per minute; ejection fraction, 0–90%; maximum, 200 different heart waves). Third, the simulative coronary arteries were composed of acrylic–silicon tubes 3 and 5 mm in inner diameter with simulative plaque inside. The simulative plaque used for stenosis was made of polypropylene material with CT attenuation ranging from –100 to –50 H. Fourth, the entire simulative left ventricle and the coronary arteries were submerged in a tank filled with water (0 H).


Figure 1
View larger version (131K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1A —Phantom system and balloon. Photographs show cardiac phantom system (A) and balloon (B) used to simulate left ventricle, attached tubes (simulated coronary arteries), and water-filled tank in which balloon is submerged. Phantom system consists of five components: driver, controller, fixed support, rubber balloon (phantom), and ECG simulator. Controller with ECG synchronizer drives balloon. Motion is achieved with four driver sequences: two speeds of fast emptying for systolic phase and fast and slow filling for diastolic phase. Balloon is filled with contrast medium to simulate contrast-enhanced heart. Acrylic tubes (mimicking coronary arteries) containing contrast medium are attached to balloon surface. Fabricated plaques are packed inside tubes to simulate stenosis. Ends of balloon are stabilized to fixed support at distance of 10 cm.

 

Figure 2
View larger version (147K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1B —Phantom system and balloon. Photographs show cardiac phantom system (A) and balloon (B) used to simulate left ventricle, attached tubes (simulated coronary arteries), and water-filled tank in which balloon is submerged. Phantom system consists of five components: driver, controller, fixed support, rubber balloon (phantom), and ECG simulator. Controller with ECG synchronizer drives balloon. Motion is achieved with four driver sequences: two speeds of fast emptying for systolic phase and fast and slow filling for diastolic phase. Balloon is filled with contrast medium to simulate contrast-enhanced heart. Acrylic tubes (mimicking coronary arteries) containing contrast medium are attached to balloon surface. Fabricated plaques are packed inside tubes to simulate stenosis. Ends of balloon are stabilized to fixed support at distance of 10 cm.

 
To investigate the effects of different vascular attenuation values and to develop an appropriate protocol for clinical use, the heart rate was set at 55 beats per minute according to the most common patient conditions in clinical cardiac examinations. Tubes in the phantom with diameters of 3 and 5 mm and stenosis of 25%, 50%, and 75% of luminal diameter were used. Fabricated plaques with approximately –60 H CT attenuation were used to simulate different degrees of stenosis in the vessels. Iopromide (Ultravist 370 mg I/mL, Bayer HealthCare) was used as the contrast medium and was prepared to enhance the simulative heart and coronary arteries according to the following four CT attenuation levels: low attenuation, 200 H; moderately low attenuation, 300 H; moderately high attenuation, 350 H; high attenu ation, 500 H.

Data Acquisition
CTA was performed with a 64-MDCT scanner (LightSpeed Volume CT, GE Healthcare). Each scan was obtained with a collimation of 64 x 0.625 mm, pitch of 0.2, and gantry rotation time of 350 milliseconds at 120 kV and 550 mAs. With the simultaneously recorded ECG data, the raw data set was retrospectively reconstructed at 5% to 95% of the R-R interval in 5% steps. Images reconstructed at 75% of the R-R interval were selected for further processing because this phase was the least affected by motion artifacts. The axial images were reconstructed with a standard kernel, field of view of 130 x 130 mm, 512 x 512 matrix size, and a slice thickness of 0.625 mm.

Data Postprocessing and Image Analysis
For further postprocessing and image analysis, all data sets were transferred from the CT scanner to an off-line workstation (D530 CMT, HP). Four indexes were used for quantitative evaluation of stenosis detection: detectability rate, cross-sectional area (CSA) of plaque, plaque attenuation, and percentage stenosis. Detectability rate was defined as the ratio between the number of image samples in which plaque was detectable and the total number of image samples. Recognized plaque successfully segmented from the vessel wall was defined as detectable. This definition of detectability differed from clinical convention. It was chosen on the basis of the actual plaque state of the phantom, which was low density and easily separated from the vessel wall. Plaque CSA was defined as the pixel numbers covered by plaque on an axial sectional image. Plaque density was defined as the average CT attenuation value covered by plaque on the axial sectional image. The best key index, percentage stenosis, which was directly relevant to the detection of luminal narrowing, was used to evaluate the accuracy of coronary stenosis detection for each attenuation level. Perc entage stenosis was defined as the ratio between plaque CSA and the CSA sum of the contrast-filled lumen and plaque on the axial sectional image. The purpose of defining per centage stenosis as an area ratio was to evaluate stenosis imaging quality synthetically in two dimensions.

Data processing was performed in the Matlab programming environment (Math Works). First, two independent cardioradiologists selected the regions of interest (ROIs) containing the target vessels. After selection of the ROIs, a threshold-adaptive method was used to calculate the thresholds for lumens and plaque segmentation according to the CT attenuation distribution in each ROI of each axial image [13]. Plaque and the contrast-filled lumen then were segmented by application of the threshold in each ROI. Finally, the measurements of areas and CT attenuation were performed with the segmentation results. The last three steps were performed automatically with a specific Matlab program designed to implement the data-processing method. The process and results of the data postprocessing are shown in Figure 2. The data-processing method had been validated with a standard phantom, and the images and results showed that the method was accurate, reproducible, and independent of observers for the calculation of percentage stenosis in a phantom [13].


Figure 3
View larger version (91K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2 —Balloon filled with contrast medium simulates beating heart and simulated vessels during stenosis detection in cardiac phantom. A, Whole axial CT image shows placement of regions of interest. Four small objects attached to balloon are four acrylic tubes simulating coronary arteries. Circles indicate simulated vessels with stenosis. B, CT image with observer input shows selection of region of interest for 5-mm vessel.C, CT image shows segmentation whereby plaque used to simulate stenosis and contrast-filled lumen are separated from vessel wall and from each other.

 

Actual plaque CSA and percentage stenosis of the phantom were defined as the reference standards in this study. The values of CSA and percentage stenosis obtained from data processing of the CTA images were compared with these standards for the investigation of cardiac imaging accuracy in coronary artery stenosis. The actual plaque CSA according to percentage stenosis values and vessel diameters of the phantom is shown in Table 1.


View this table:
[in this window]
[in a new window]

  		TABLE 1: Summary of Plaque Cross-Sectional Area Accuracy Values for Vessel Attenuation Groups

 

Statistical Analysis
Statistical results of the data complying with a normal distribution were expressed as mean ± SD. For variable comparisons between two groups, independent samples Student's t tests were performed. Comparison among groups was performed with one-way analysis of variance. Differences of sample rates were examined with a chi-square test. All the tests were performed with SPSS software (version 11.0, SPSS). Statistical significance was accepted at p < 0.05. The authors had full access to and took responsibility for the integrity of the data.


Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
All examinations were performed without technical problems, and the image quality was good for data analysis in all cases. A total of 1,080 axial image samples containing the simulative vessels in two sizes and three degrees of stenosis at four attenuation levels were processed and assessed for detectability rate, plaque CSA, plaque attenuation, and percentage stenosis.

Calculation of Detectability Rate
The results showed that intracoronary attenuation not only affected the degree of stenosis but also sometimes did not allow detection of small plaques; thus the detectability rate of plaque was calculated first. All stenosis was detected successfully in larger vessels (5 mm in diameter), and 75% stenosis was detected successfully in smaller vessels (3 mm in diameter). For 50% stenosis in smaller vessels, the detectability rate at 500 H was significantly lower than for the other three groups. For 25% stenosis in smaller vessels, the difference in the detectability rate was statistically significant among the four attenuation groups (p < 0.05) (Table 2).


View this table:
[in this window]
[in a new window]

  		TABLE 2: Summary of Detectability Rates for Vessel Attenuation Groups

 

Measurement of Plaque CSA
In larger vessels, the measured plaque CSA differed significantly between the 200-H group and the 500-H group (p < 0.05). A larger area than standard was found in the 200-H group, and a smaller area was found in the 500-H group. Meanwhile, the SD of the CSA measurements in the 500-H group was much smaller than that of the 200-H group, reflecting more reproducible image quality in the high-attenuation group. There was no significant difference in plaque CSA measurements between the 300- and 350-H groups (p > 0.05). CSA measurements in these two groups were more accurate and closer to the standard area than those of the other two groups.

For smaller vessels, the measured CSA in the 500-H group was significantly smaller than the standard area. This difference reached more than 50% of the standard area in evaluation of image samples with 50% stenosis. The measured plaque CSA in the 200- to 350-H groups was larger than the standard area. For 25% and 50% stenosis, there was no significant difference among the three groups in regard to plaque CSA (p > 0.05). Table 1 summarizes the attenuation groups and results of measured plaque CSA.

Measurement of Plaque Density
For 50% and 75% stenosis of larger vessels, the plaque density values of the 500-H group were significantly lower than those of the other three groups and closer to the true CT attenuation of plaque. For all degrees of stenosis in smaller vessels and 25% stenosis in larger vessels, plaque density varied positively with vascular attenuation and negatively with degree of stenosis. Table 3 summarizes the attenuation protocols and results of average plaque density.


View this table:
[in this window]
[in a new window]

  		TABLE 3: Summary of Average Plaque Density for Vessel Attenuation Groups

 

Evaluation of Percentage Stenosis
For 50% and 75% stenosis in larger vessels, slight overestimation of degree of stenosis occurred in the 200-H group compared with the other three groups, but the difference was not statistically significant. The measured percentages of stenosis in the 300- to 500-H groups were more accurate, and the shifts from the stenosis standard were all smaller than 5%. For 50% and 75% stenosis in smaller vessels, slight overestimation of degree of stenosis also occurred in the 200-H group compared with the 300- and 350-H groups, but the difference was not statistically significant. Significant underestimation of percentage stenosis was revealed in the 500-H group compared with the other three groups (p < 0.05).

For 25% stenosis in larger vessels, overestimation of degree of stenosis occurred in the 200-H group compared with the other three groups, and there were statistically significant differences between the 200-H group and the other three groups (p < 0.05). For 25% stenosis in smaller vessels, not all stenosis was detected on the images in the 500-H group, the high-attenuation protocol. The accuracy of measured percentage stenosis is shown in Figure 3A, 3B, 3C for different degrees of stenosis; detailed data are shown in Table 4.


Figure 4
View larger version (16K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 3A —Accuracy of measured percentage stenosis. Values are means; error bars indicate mean ± 0.5 SD; light gray, 3-mm vessel diameter; dark gray, 5-mm vessel diameter. Graph shows effects of attenuation protocols of cardiac CT angiography (CTA) on accuracy of detection of 25% stenosis detection. High-attenuation (500 H) protocol resulted in underestimation of stenosis. Stenosis was especially difficult to detect in smaller vessels. Low-attenuation (200 H) protocol led to overestimation of stenosis detection within clinically acceptable limits, but imaging quality was much less reproducible than in other protocols.

 

Figure 5
View larger version (21K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 3B —Accuracy of measured percentage stenosis. Values are means; error bars indicate mean ± 0.5 SD; light gray, 3-mm vessel diameter; dark gray, 5-mm vessel diameter. Graph shows effects of attenuation protocols of cardiac CTA on accuracy of detection of 50% stenosis. High-attenuation (500 H) protocol led to significant underestimation of stenosis for smaller vessels but was accurate in depicting larger vessels. Low-attenuation (200 H) protocol led to overestimation of stenosis within clinically acceptable limits, but imaging quality was not as reproducible as in other groups.

 

Figure 6
View larger version (21K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 3C —Accuracy of measured percentage stenosis. Values are means; error bars indicate mean ± 0.5 SD; light gray, 3-mm vessel diameter; dark gray, 5-mm vessel diameter. Graph shows effects of attenuation protocols of cardiac CTA imaging on accuracy of detection of 75% stenosis. High-attenuation (500 H) protocol led to significant underestimation of stenosis in smaller vessels but was accurate in depicting larger vessels. Low-attenuation (200 H) protocol led to slight overestimation of stenosis. Imaging accuracy and stability did not differ significantly from moderate-attenuation protocols (300 and 350 H).

 

View this table:
[in this window]
[in a new window]

  		TABLE 4: Summary of Measured Accuracy of Percentage Stenosis

 


Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Contrast-enhanced coronary CTA has been widely used for the noninvasive detection of coronary disease. Thus the use of contrast media is becoming a popular research topic in coronary CTA technology. How vascular attenuation affects image quality and the accuracy of detection of coronary stenosis, the extent of stenosis, and the most appropriate enhancement level for coronary CTA are frequently asked questions. Previous studies [9, 14] have shown that higher intracoronary attenuation facilitates reliable visualization of the coronary arteries. Other reports [10] have suggested that high contrast enhancement can interfere with coronary calcifications. In this study, we performed mass repetitive experiments using identical scan protocols with a pulsating phantom of standard size to investigate the effects of vascular attenuation on the accuracy of detection of coronary stenosis in coronary CTA with 64-MDCT.

Previous studies of in vivo quantitative assessments of coronary stenosis with MDCT have shown a significant correlation with findings at intravascular sonography but only moderate accuracy in quantification. Moselewski [15] found that mean luminal area and mean plaque area were slightly but significantly overestimated with 16-MDCT. Leber et al. [3] reported underestimation with 64-MDCT compared with intravascular sonography. Those results were explained by use of different protocols for different types of MDCT; different settings for the display of MDCT images; failure of MDCT to depict plaque in some cross sections; and, especially, interobserver variability. In our study, quantification was based on the pulsating cardiac phantom, in which plaque size and shape were regular and plaque density was uniform. In addition, the threshold-adaptive method was specifically designed for plaque and lumen segmentation in this phantom. The only observer input was identification of ROIs; thus differences in observer ability did not affect the quantitative results.


Figure 7
View larger version (53K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 4A —Phantom with simulated vessels with 50% stenosis. Results show low attenuation decreased contrast-to-noise ratio and caused overestimation of stenosis, especially in small vessels. High attenuation enhanced contrast-to-noise ratio but also caused severe partial volume effect. Thus larger stenosis in larger vessel was clearly depicted but stenosis was underestimated in smaller vessels. CT attenuation of 350 H is optimal for accurate quantification of stenosis in larger and smaller vessels. CT scan shows 3- and 5-mm coronary arteries under low-attenuation (200 H) protocol.

 


Figure 8
View larger version (52K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 4B —Phantom with simulated vessels with 50% stenosis. Results show low attenuation decreased contrast-to-noise ratio and caused overestimation of stenosis, especially in small vessels. High attenuation enhanced contrast-to-noise ratio but also caused severe partial volume effect. Thus larger stenosis in larger vessel was clearly depicted but stenosis was underestimated in smaller vessels. CT attenuation of 350 H is optimal for accurate quantification of stenosis in larger and smaller vessels. CT scan shows 3- and 5-mm coronary arteries under moderately-high-attenuation (350 H) protocol.

 


Figure 9
View larger version (45K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 4C —Phantom with simulated vessels with 50% stenosis. Results show low attenuation decreased contrast-to-noise ratio and caused overestimation of stenosis, especially in small vessels. High attenuation enhanced contrast-to-noise ratio but also caused severe partial volume effect. Thus larger stenosis in larger vessel was clearly depicted but stenosis was underestimated in smaller vessels. CT attenuation of 350 H is optimal for accurate quantification of stenosis in larger and smaller vessels. CT scan shows 3- and 5-mm coronary arteries under high-attenuation (500 H) protocol.

 
The results of this study showed that high attenuation (500 H) resulted in accurate stenosis detection and stable imaging quality only for larger vessels ({approx} 5 mm in diameter) and for greater degrees of stenosis (Fig. 4C) because of a higher contrast-to-noise ratio, and fewer pixels were affected by the partial volume effect compared with that of plaque CSA. However, with high attenuation for smaller vessels (3 mm in diameter), the contrast-filled lumen showed a severe partial volume effect on plaque. Thus the CT attenuation of plaque increased on a large scale, and plaque CSA and percentage stenosis were significantly underestimated. Especially for the low degree of stenosis, almost no plaque was differentiated from the vessel walls on the images (Fig. 5C). A similar result occurred with a low degree of stenosis in larger vessels, which may be explained by the fact that a smaller area was more easily affected by the partial volume effect because the ratio between the pixels affected by the partial volume effect and pixels of total plaque CSA was larger.


Figure 12
View larger version (106K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 5C —Phantom with simulated vessels with 25% stenosis. Results show low attenuation decreased contrast-to-noise ratio and thus caused overestimation of stenosis, especially in small vessels. High attenuation enhanced contrast-to-noise ratio but also caused severe partial volume effect on small stenosis, especially in small vessels. Intracoronary CT attenuation of 350 H was optimal for accurate quantification of coronary stenosis for larger and smaller vessels. CT scan shows 3- and 5-mm coronary arteries under high-attenuation (500 H) protocol.

 
Overestimation of stenosis always occurred with use of contrast medium with low attenuation ({approx} 200 H), but the degree of overestimation was within the acceptable limits for clinical use; the largest average difference was 11.1% (36.1% – 25% = 11.1%). In addition, the effect of low contrast attenuation on plaque density was smaller than that of high attenuation. However, the results in the low-attenuation group did not differ significantly from the results in the moderate-attenuation groups. On the basis of the results of statistical analysis, we conclude that at low attenuation, image quality was not stable enough because the large deviation of results due to decreased contrast-to-noise ratio resulted in reduced sharpness of vessel contours [9] (Figs. 4A and 5A).


Figure 10
View larger version (142K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 5A —Phantom with simulated vessels with 25% stenosis. Results show low attenuation decreased contrast-to-noise ratio and thus caused overestimation of stenosis, especially in small vessels. High attenuation enhanced contrast-to-noise ratio but also caused severe partial volume effect on small stenosis, especially in small vessels. Intracoronary CT attenuation of 350 H was optimal for accurate quantification of coronary stenosis for larger and smaller vessels. CT scan shows 3- and 5-mm coronary arteries under low-attenuation (200 H) protocol.

 
Stenosis detection with a moderately-high-attenuation protocol ({approx} 350 H) was the most accurate and closest to the standard degree of stenosis among the four groups (largest difference of percentage stenosis < 6%), and the image quality was the most stable in most cases (Figs. 4B and 5B). This finding occurred because vessel enhancement with appropriate attenuation does not render plaque less detectable owing to severe partial volume effect. At the same time, this method avoids low sharpness of the contours between plaque, contrast-filled lumen, and vessel walls caused by a reduced contrast-to-noise ratio with poor enhancement of vessel lumens.


Figure 11
View larger version (126K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 5B —Phantom with simulated vessels with 25% stenosis. Results show low attenuation decreased contrast-to-noise ratio and thus caused overestimation of stenosis, especially in small vessels. High attenuation enhanced contrast-to-noise ratio but also caused severe partial volume effect on small stenosis, especially in small vessels. Intracoronary CT attenuation of 350 H was optimal for accurate quantification of coronary stenosis for larger and smaller vessels. CT scan shows 3- and 5-mm coronary arteries under moderately-high-attenuation (350 H) protocol.

 

It is not worth recommending low vessel attenuation in coronary CTA examinations. However, as the number of patients undergoing CT increases, more patients are affected by adverse events due to contrast media, especially contrast-induced nephropathy. Previous studies [16, 17] have shown that volume of contrast material is an independent predictor of deterioration in renal function among patients with chronic kidney disease (serum creatinine concentration ≥ 1.8 or 2.0 mg/dL). Experts in the field and members of the European Society of Urogenital Radiology agree that the risk of contrast-induced nephropathy increases with increasing dose of contrast material [18]. One recommended method for preventing contrast-induced nephropathy is to control the volume of contrast medium, although the optimal threshold dose has not yet been established [19]. Because the accuracy of detection under low vascular attenuation is within acceptable limits for clinical use, controlled volume of contrast medium may be used when coronary CTA is necessary in the care of patients with mild renal dysfunction to reduce risk of contrast-induced nephropathy.

We used only the standard kernel in this study because it is most often used in cardiac CTA in the clinic. Reports [20] on the effect of different convolution kernels on the measurement of vascular diameter have suggested that variations in convolution kernels also may play a key role in the accuracy of stenosis measurement. Further studies on the effect of different convolution kernels combined with attenuation levels should be performed.

This study had limitations. First, the step length of vascular attenuation was somewhat large, and thus the data were not sufficient for drawing the relation curve between the accuracy of detection of stenosis and attenuation levels. Second, because of phantom limitations, this study did not include vessels smaller than 2 mm in diameter or those with calcifications. Third, this study was an in vitro assessment of coronary stenosis. Plaque shape was always regular, a condition different from plaque conditions in vivo; thus plaque density did not fully reflect the in vivo situation.

The optimal vascular attenuation for detection of coronary artery stenosis on CTA with 64-MDCT is approximately 350 H. Contrast enhancement at relatively lower attenuation ({approx} 200 H) also is clinically acceptable, although slight overestimation in stenosis depiction can occur. Superhigh attenuation (> 500 H) decreases the detectability of stenosis in smaller vessels and is not to be used in coronary CTA examinations.


References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Nikolaou K, Flohr T, Knez A, et al. Advances in cardiac CT imaging: 64-slice scanner. Int J Cardiovasc Imaging2004; 20:535 –540[CrossRef][Medline]
  2. Achenbach S, Ulzheimer S, Baum U, et al. Noninvasive coronary angiography by retrospectively ECG-gated multislice spiral CT. Circulation 2000;102 :2823 –2828[Abstract/Free Full Text]
  3. Leber AW, Knez A, von Ziegler F, et al. Quantification of obstructive and nonobstructive coronary lesions by 64-slice computed tomography: a comparative study with quantitative coronary angiography and intravascular ultrasound. J Am Coll Cardiol2005; 46:147 –154[Abstract/Free Full Text]
  4. Yamashita Y, Komohara Y, Takahashi M, et al. Abdominal helical CT: evaluation of optimal doses of intravenous contrast material—a prospective randomized study. Radiology2000; 216:718 –723[Abstract/Free Full Text]
  5. Awai K, Takada K, Onishi H, Hori S. Aortic and hepatic enhancement and tumor-to-liver contrast: analysis of the effect of different concentrations of contrast material at multi-detector row helical CT. Radiology 2002;224 : 757–763[Abstract/Free Full Text]
  6. Kim T, Murakami T, Takahashi S, et al. Effects of injection rates of contrast material on arterial phase hepatic CT. AJR1998; 171:429 –432[Abstract/Free Full Text]
  7. Awai K, Hiraishi K, Hori S. Effect of injection duration and rate on aortic peak time and peak enhancement at dynamic CT using contrast injection protocol with dose tailored to patient weight. Radiology 2004;230 : 142–150[Abstract/Free Full Text]
  8. Awai K, Inoue M, Yagyu Y, et al. Moderate versus high concentration of contrast material for aortic and hepatic enhancement and tumor-to-liver contrast at multi–detector row CT. Radiology2004; 233:682 –688[Abstract/Free Full Text]
  9. Cademartiri F, Mollet NR, Lemos PA, et al. Higher intracoronary attenuation improves diagnostic accuracy in MDCT coronary angiography. AJR 2006; 187:1115; [web]W430–W433
  10. Becker CR, Hong C, Knez A, et al. Optimal contrast application for cardiac 4-detector-row computed tomography. Invest Radiol 2003; 38:690 –694[Medline]
  11. Horiguchi J, Shen Y, Akiyama Y, et al. Electron beam CT versus 16-MDCT on the variability of repeated coronary artery calcium measurements in a variable heart rate phantom. AJR 2005;185 :995 –1000[Abstract/Free Full Text]
  12. Fei X, Du X, Li P, Liao J, Shen Y, Li K. Effect of dose-reduced scan protocols on cardiac coronary image quality with 64-row MDCT: a cardiac phantom study. Eur J Radiol 2007;Aug 30 [Epub ahead of print]
  13. Fei X, Li K, Du X, et al. An objective evaluation method designed for pulsating cardiac coronary phantom based on 64-row detector CT imaging [in Chinese]. Inf Med Equip 2007;22 : 3–6
  14. Roos JE, Desbiolles LM, Weishaupt D, et al. Multi-detector row CT: effect of iodine dose reduction on hepatic and vascular enhancement. Rofo 2004; 176:556 –563[Medline]
  15. Moselewski F. Comparison of measurement of cross-sectional coronary atherosclerotic plaque and vessel areas by 16-slice multidetector computed tomography versus intravascular ultrasound. Am J Cardiol 2004; 94:1294 –1297[CrossRef][Medline]
  16. Gruberg L, Mintz GS, Mehran R, et al. The prognostic implications of further renal function deterioration within 48 h of interventional coronary procedures in patients with pre-existent chronic renal insufficiency. J Am Coll Cardiol 2000;36 :1542 –1548[Abstract/Free Full Text]
  17. Taliercio CP, Vlietstra RE, Fisher LD, Burnett JC. Risks for renal dysfunction with cardiac angiography. Ann Intern Med1986; 104:501 –504[Abstract/Free Full Text]
  18. Morcos SK, Thomsen HS, Webb JA. Contrast-media-induced nephrotoxicity: a consensus report. Contrast Media Safety Committee, European Society of Urogenital Radiology (ESUR). Eur Radiol1999; 9:1602 –1613[CrossRef][Medline]
  19. Stacul F. Managing the risk associated with use of contrast media for computed tomography. Eur J Radiol2007; 62:33 –37[CrossRef]
  20. Suzuki S, Furui S, Kaminaga T, et al. Measurement of vascular diameter in vitro by automated software for CT angiography: effects of inner diameter, density of contrast medium, and convolution kernel. AJR 2004; 182:1313 –1317[Abstract/Free Full Text]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 Am. J. Roentgenol.Home page
P. T. Johnson, H. K. Pannu, and E. K. Fishman
IV Contrast Infusion for Coronary Artery CT Angiography: Literature Review and Results of a Nationwide Survey
Am. J. Roentgenol., May 1, 2009; 192(5): W214 - W221.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Fei, X.
Right arrow Articles by Li, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Fei, X.
Right arrow Articles by Li, K.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
Hotlight (NEW!)
Right arrow
What's Hotlight?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS