(Radiology. 2000;217:494-500.)
© RSNA, 2000
CT Angiographic Measurement of the Carotid Artery: Optimizing Visualization by Manipulating Window and Level Settings and Contrast Material Attenuation1
Yi Liu, MD,
Kenneth D. Hopper, MD,
David T. Mauger, PhD and
Kimberly A. Addis, MD
1 From the Departments of Radiology (Y.L., K.D.H.) and Health Evaluation Sciences (D.T.M.) and the College of Medicine (K.A.A.), Penn State University, PO Box 850, Hershey, PA 17033; and the Department of Radiology, Hua Dong Hospital, Shanghai, China (Y.L.). Received May 18, 1999; revision requested July 14; revision received January 24, 2000; accepted February 21. Address correspondence to K.D.H. (e-mail: khopper@psu.edu).
 |
ABSTRACT
|
|---|
PURPOSE: To evaluate a broad range of window and level settings for various contrast material attenuation coefficients and degrees of vascular stenosis to obtain the most accurate computed tomographic (CT) angiographic measurements.
MATERIALS AND METHODS: A total of 25,480 measurements were made transversely (perpendicular to the lumen) and by means of maximum intensity projection (MIP) in a phantom with stenoses of 0%–100%, contrast material with attenuation coefficients of 150–350 HU, and 14 window and 13 level settings. Edge definition was also evaluated.
RESULTS: There was an inherent relationship between the contrast material attenuation coefficient and the optimal window and level settings in the measurement of stenoses at both transverse and MIP CT angiography. This relationship between the contrast material attenuation coefficient D and the optimal settings for window W and level L was represented by the following simple equations: W/D = [-2 x (L/D)] + 1.3, where -0.2 < L/D < 0.5, and W/D = [3.3 x (L/D)] - 1.3, where 0.5 < L/D < 1.0. With a vascular contrast material attenuation coefficient of 250–350 HU, the best transverse and MIP display settings for the window and level were 96 and 150 HU, respectively.
CONCLUSION: The use of optimized window and level settings at CT angiography reduces measurement variability.
Index terms: Carotid arteries, angiography, 172.129116 • Carotid arteries, CT, 172.129116 • Carotid arteries, stenosis or obstruction, 172.721 • Computed tomography (CT), maximum intensity projection, 172.129119 • Contrast media, experimental studies • Phantoms
|
INTRODUCTION
|
|---|
Computed tomographic (CT) angiography (especially with the introduction of multisection CT) can reliably depict arterial structures with the use of only intravenously administered contrast material (1–11). However, there is disagreement in the radiology literature about the optimal window and level settings for use in displaying transverse (perpendicular to the vascular lumen) and maximum intensity projection (MIP) CT angiographic images (12–21).
Claves et al (12) found that the accuracy of CT angiography was significantly affected by the attenuation coefficient of contrast material within a particular vessel. After extensive study, Dix et al (13) found that the highest precision with CT angiographic measurement was obtained by using magnified transverse images with a window setting of zero and a level setting that was halfway between the luminal contrast material and vascular wall attenuation coefficients. However, Lev et al (14) found that multiple window and level settings resulted in accurate measurements. They found that a level setting of full width at half maximum between the true Hounsfield attenuation of the object being measured and the background attenuation and that a window setting at least three times the level value were optimal.
Because of this apparent discrepancy in the radiology literature about the optimal window and level settings for use in obtaining the most accurate CT angiographic measurements, we designed a comprehensive study to evaluate a broad range of window and level settings for a variety of contrast material attenuation coefficients and degrees of vascular stenosis.
|
MATERIALS AND METHODS
|
|---|
A phantom was designed specifically for this study (Fig 1). The native luminal diameter was 8 mm. One-centimeter-long 0%, 25.0%, 50.0%, 75.0%, 87.5%, 93.8%, and 100% stenoses were placed in the center of the phantom vessels. The angle of inclination from the normal lumen to the stenosis was a constant 75°. The stenosis was concentric and centered on the normal lumen.
View larger version (108K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1. Photograph shows the phantom placed in the scanner for imaging. Appropriate stenotic vessels are filled with contrast material and are oriented along the long axis of the table.
|
|
Five Plexiglas containers each contained 300 mL of contrast material (iohexol; Nycomed Amersham, Oslo, Norway) at room temperature that was carefully diluted to have an attenuation coefficient of 150, 200, 250, 300, or 350 HU. Subsequently, the phantom vessels were filled with contrast material having each attenuation coefficient. The phantom was scanned with a 2-mm section thickness, and images were reconstructed with a 1.0-mm interval. We used 175 mA, 120 kV, a pitch of 1.0, a standard interpolator, and standard algorithms (PQ 5000; Picker International, Highland Heights, Ohio). The vessels were scanned in an orientation along the long axis of the scanner and parallel to the z axis. Between the use of contrast material with each different attenuation coefficient, the phantom vessels were carefully flushed with distilled water and dried. Prior to acquisition of the study image, preliminary images were obtained to ensure that all air bubbles had been removed.
Each of the five data sets corresponding to the different contrast material attenuation coefficients were evaluated by means of transverse (perpendicular to the lumen) and MIP CT angiographic methods with the use of 14 equally spaced window settings ranging from zero to three times the contrast material attenuation coefficient and 13 equally spaced levels ranging from -0.1 to 1.0 times the contrast material attenuation coefficient. Each vessel was evaluated at two sites: the midportion of its proximal normal 8-mm lumen and the middle of the stenosis. This procedure yielded a total of 25,480 measurements (Table 1).
The data set was loaded onto a workstation (Voxel Q; Picker International). A single experienced investigator (Y.L.) who was blinded to the actual vascular diameter performed all of the measurements to an accuracy of 0.1-mm using electronic calipers. A magnification of x8 was used on every image. At each site evaluated, three measurements were obtained and averaged. In addition to luminal diameter measurements, the sharpness of the vascular edges was also evaluated at every measured site with respect to the penumbra (halo) depicted about the lumen (Fig 2). The absolute error was calculated for each measurement as the difference between the measured and known vascular diameters (within 1 µm), as determined by the manufacturer.
View larger version (98K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2. Transverse CT angiograms of a 4-mm-diameter vessel depict four types of penumbra (halo): Grade A, no halo or blurring; Grade B, minimal halo (blurring) about the outside edge of the opacified lumen, which is still sharply defined; Grade C, marked blurring between the halo and outside luminal edge, which makes measurement difficult; Grade D, sharp outside vascular edge, and halo or blurriness appears just inside this edge. Variability in measurement (and halo) was caused by using widely different window and level settings. T = diameter.
|
|
For each combination of window or level setting and contrast material attenuation coefficient, the mean and SD of the absolute error were calculated. Simultaneously plotting the mean absolute error as a function of both window or level setting and contrast material attenuation coefficient allowed us to determine the window and level settings that provided the most accurate measurements (ie, small absolute error). Equations corresponding to the regions of smallest error were estimated by means of visual assessment. Similar results were obtained by means of response surface regression. This analysis was performed separately for measurements obtained with the transverse and MIP CT angiographic methods. All analyses were performed by using SAS and S-Plus (SAS, Cary, NC) statistical software packages.
|
RESULTS
|
|---|
The smaller the actual luminal diameter, the greater the potential measurement error (Figs 3, 4). This was true with both transverse and MIP CT angiographic methods. The larger the actual vascular diameter, the less the percentage measurement variability with both methods, although this decreased measurement variability was more pronounced with the transverse method. With larger vessels, MIP tended to cause undermeasurement of the actual luminal diameter. With high degrees (87.5%–93.8%) of stenosis, there was tremendous variability in the percentage measurement in most but not all of the contrast material attenuation coefficients. There was little variation with high degrees of stenosis and contrast material attenuation coefficients of 200 (transverse and MIP), 250 (transverse), and 300 (transverse) HU.
View larger version (54K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3. Line graphs depict mean absolute error versus percentage of stenosis (left), window width (middle), or imaging level (right) for transverse (top) and MIP (bottom) images. The smaller the actual luminal diameter, the greater the percentage of measurement variability. In larger vessels, MIP tended to cause underestimation of the diameter. There was little measurement variability with a contrast material attenuation coefficient of at least 200 HU.
|
|
View larger version (55K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4. Transverse (top) and MIP (bottom) CT angiograms with window and level settings optimized for the contrast material attenuation coefficient show four vessels with central stenotic luminal diameters of 1, 2, 4, and 8 mm. Compared with actual measurements, diameter measurements on images varied little in vessels 2 mm or smaller but varied considerably in those 1 mm or smaller. T = diameter.
|
|
For each contrast material attenuation coefficient evaluated, we found a variety of window and level settings that could be used to accurately display the lumen. By using ratios of a window or level setting to the contrast material attenuation coefficient and by displaying the absolute mean error for both the window or level setting and contrast material attenuation coefficient combination simultaneously, the best window and level settings were easily visualized (Figs 5–7). The window and level setting was optimized for all contrast material attenuation coefficients from 150 to 350 HU (Fig 5). When displayed in this way, the darker areas represented greater measurement error; the whiter areas, less measurement error. The acceptable window or level setting and contrast material attenuation coefficient combinations formed two intersecting lines. Mathematic evaluations of these functions were used to create the following equations that represented these lines: W/D = [-2 x (L/D)] + 1.3, where -0.2 < L/D < 0.5, and W/D = [3.3 x (L/D)] - 1.3, where 0.5 < L/D < 1.0. D was the contrast material attenuation coefficient, L was the level setting, and W was the window setting.
View larger version (63K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5a. Plots show that when the mean absolute error for window or level setting and contrast material attenuation coefficient ratio at (a) transverse and (b) MIP imaging are displayed simultaneously, the best window and level settings are easily visualized. Darker areas represent more measurement error; whiter areas, less measurement error. Acceptable window or level setting and contrast material attenuation coefficient ratios form two intersecting lines.
|
|
View larger version (63K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5b. Plots show that when the mean absolute error for window or level setting and contrast material attenuation coefficient ratio at (a) transverse and (b) MIP imaging are displayed simultaneously, the best window and level settings are easily visualized. Darker areas represent more measurement error; whiter areas, less measurement error. Acceptable window or level setting and contrast material attenuation coefficient ratios form two intersecting lines.
|
|
View larger version (74K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6. Transverse (top) and MIP (bottom) CT angiograms show a 4-mm-diameter, 1-cm-long stenosis. A and B, Images depict the vessel sharply with a contrast material attenuation coefficient of 300 HU and the crossover point of the two lines for the optimal window or level setting and contrast material attenuation coefficient (window setting, 90 HU; level setting, 150 HU). C and D, Images show that equivalent measurement results were obtained by using window (390 HU) and level (0 HU) settings from the top of the two intersecting lines for the optimal window or level setting and contrast material attenuation coefficient. E and F, Images show that use of a center point chosen from along the bottom optimal line for window or level setting and contrast material attenuation coefficient also results in accurate measurements (window setting, 501 HU; level setting, 270 HU). T = diameter.
|
|
View larger version (62K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7a. (a) Transverse and (b) MIP images show the same 4.0-mm vessel as in Figure 6. Top left images in a and b were obtained with optimized window (W) and level (L) settings at the crossover point, with a contrast material attenuation coefficient of 300 HU. Subsequent images were obtained in the four quadrants in Figure 5 with suboptimal window and level settings for this contrast material attenuation coefficient. With this 4.0-mm lumen, the negative effect on measurement accuracy is dramatic. T = diameter.
|
|
View larger version (60K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7b. (a) Transverse and (b) MIP images show the same 4.0-mm vessel as in Figure 6. Top left images in a and b were obtained with optimized window (W) and level (L) settings at the crossover point, with a contrast material attenuation coefficient of 300 HU. Subsequent images were obtained in the four quadrants in Figure 5 with suboptimal window and level settings for this contrast material attenuation coefficient. With this 4.0-mm lumen, the negative effect on measurement accuracy is dramatic. T = diameter.
|
|
The optimal window and level settings (crossover point) for both techniques and all contrast material attenuation coefficients were also determined (Table 2). The sharpness of the vascular edges had a considerable effect on measurement accuracy. The greater the contrast material attenuation coefficient, the greater the proportion of blurred edges (grade C margin), which was slightly greater with MIP images versus transverse images. Also, the proportion of blurred edges was much greater with more severe stenoses and was most marked in vessels with 87.5%–93.8% stenoses. However, the greatest cause of vascular blurring was the selection of suboptimal window and level settings for a particular contrast material attenuation coefficient (Fig 8).
View larger version (60K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8. Bar graphs show the proportion of luminal edge halo as a function of the contrast material attenuation coefficient, display method, percentage of vascular stenosis, and optimized window and level settings. Color for halo grades are as follows: green, A; blue, B; red, C; and black, D. With a grade C halo, the luminal edge for measurement is the worst and is dramatically reduced when optimal window and level settings are used.
|
|
|
DISCUSSION
|
|---|
In a meticulous phantom study of short (1–5-mm) stenoses, Claves et al (12) found that the concentration of contrast material with a particular attenuation coefficient within a vascular lumen significantly affected the accuracy of measurement. Contrast material attenuation coefficients of less than 150 HU or greater than 250 HU caused increased measurement inaccuracy. Claves et al also evaluated the effect of ionic versus nonionic contrast materials and obtained the most accurate measurements with ionic and nonionic contrast materials with attenuation coefficients of 150 and 150–200 HU, respectively. However, Claves et al used constant window and level settings and evaluated only the differences in the luminal contrast material attenuation coefficient. However, in clinical practice, the window and level display settings are varied more easily than are the intravascular contrast material attenuation. Although the use of a power-injector to administer the bolus should ensure a high contrast material attenuation coefficient within the lumen to be measured, there is certain to be variability in the luminal contrast material attenuation coefficient between patients and vessels.
Dix et al (13) also used a phantom with stenoses ranging from 50% to 95%. Unlike Claves et al, who evaluated only transverse images, Dix et al evaluated transverse, MIP, and shaded-surface images obtained with a variety of window and level settings. They found that systematic error was produced by changing the level setting from a value of approximately halfway between the luminal contrast and vascular wall attenuation coefficients. They found that random error was introduced with the use of window settings greater than zero. Dix et al concluded that CT angiography was highly accurate, especially when transverse images were used with a window setting of zero and a level setting halfway between the luminal contrast and vascular wall attenuation coefficients.
Lev et al (14) used a phantom with 0.5-, 1.0-, and 1.5-mm vessels. They varied the scanning pitch from 1 to 2, the collimation from 1 to 3 mm, and the reconstruction algorithms from standard to high spatial resolution. Using five display level settings and four window settings, they evaluated measurement accuracy with their phantom. They found that the most precise diameter measurements were obtained with a level setting of full width at half maximum between the true Hounsfield attenuation of the object being measured and the background attenuation and with a window setting at least three times the level value. With a optimal window and level settings, he obtained a measurement error of ±0.1 mm at CT angiography. Like Dix et al, Lev et al used contrast material with a constant attenuation coefficient.
Diederichs et al (15) evaluated CT angiographic MIP in both the transverse and longitudinal planes using a cone phantom (117 HU) immersed in water. Using a binary window width of 2, they evaluated display levels of 20%, 50%, and 80% of the difference in the contrast material attenuation coefficient. With a transverse orientation and a window width of 2, they found that the best level was 20% of the attenuation coefficient difference. With a longitudinal orientation, however, they found that the optimal level was halfway between the attenuation coefficient of the cone and water. The findings of Diederichs et al are similar to ours except that the attenuation coefficient of the cone and water are not useful representations of soft tissue and a well-enhanced opacified vessel. In addition, they did not vary the window width through a broad spectrum of display levels.
Vandermeulen et al (16) found that no single optimal window or level setting could be used to reliably display angiograms. Rather, they recommend the use of a display level setting equal to the full width at half maximum for the selected section thickness. These results are similar to those of Lev et al (14) except that multiple window settings were not evaluated, and Vandermeulen et al apparently used the full width at half maximum for the selected section thickness rather than the full width at half maximum between the luminal contrast attenuation coefficient and background attenuation coefficient.
Unlike the majority of diagnostic CT imaging methods, CT angiography relies on the display of spatial resolution between a structure of high density and another structure (vascular wall and surrounding tissues) of generally much lower density. The interface between these two structures with widely varying densities creates the potential for blurring, depending on the level and window settings chosen. When a wide window width is chosen, vascular voluming effects cause an increased and widened penumbra (halo) between the two structures. When a very narrow window width is used, this edge blurring can be minimized. These choices account for our data that were obtained by crossing the two lines for ideal window and level settings near a window width of zero. Our data, however, also demonstrate that the edge blurring with the use of wider window widths can be at least partially compensated for by changing the placement of the display level. In addition, even with a window setting of zero or nearly zero, a suboptimal display level setting for a given contrast material attenuation coefficient can have disastrous results. The use of this crossover point of the two selection lines for optimal widow and level settings provides the most accurate measurements regardless of the measured luminal diameter; although, the smaller the lumen is, the more sensitive the measurements are to suboptimal selection of the window and level setting.
This study was an evaluation of the accuracy of CT angiographic measurement as a function of the window setting, level setting, and contrast material attenuation coefficient. Our results demonstrate that while the type of projection (transverse, MIP) and contrast material attenuation coefficient somewhat affect measurement accuracy in vessels larger than 1.0 mm in diameter, the most important determinant of measurement accuracy is the selection of the optimal window and level settings. Regardless of the contrast material attenuation coefficient, injection rate, and scanning delay, considerable variability can be expected in vascular contrast material attenuation coefficients from patient to patient and from vessel to vessel. As a result, the best window and level settings to use in the evaluation of different patients or vessels will vary and could be individually optimized to obtain the most accurate measurements. Analysis of our data have yielded straightforward easy-to-use equations for the calculation of optimal window and level settings for each vessel and/or patient examined at CT angiography. We have also calculated the optimal window and level settings (crossover point) for a variety of contrast material attenuation coefficients (Table 2).
There are two shortcomings of this study. First, all of the measurements were obtained by one experienced investigator (Y.L.). While this approach is ideal, the lack of measurements by multiple investigators prevents the evaluation of interobserver variability. However, this evaluation was not possible in this study because of the large number of measurements performed (n = 25,480). A second deficiency is that this is a phantom rather than human (clinical) study. However, this type of study (multiple scanning of vessels with stenoses verified to an accuracy of 1 µm) is not feasible in vivo. The use of a phantom provides the accuracy of known vascular diameters, the ability to repeat scanning, and the use of varying contrast material attenuation coefficients.Practical applications: The results of this study are directly applicable to clinical CT angiography performed with transverse (perpendicular to the lumen) and MIP display methods. The optimal window and level display settings have been determined for all degrees of stenoses and luminal contrast material attenuation coefficients. For vessels with excellent contrast material opacification (250– 350 HU), the best transverse and MIP display settings are a window setting of 96 HU and a level setting of 150 HU. For other contrast material attenuation coefficients, easy-to-use equations and a table are provided for use in the determination of the best window and level settings. The use of optimized window and level settings at CT angiography reduces measurement variability.
|
FOOTNOTES
|
|---|
Abbreviation: MIP = maximum intensity projection
Author contributions: Guarantor of integrity of entire study, K.D.H.; study concepts and design, K.D.H.; definition of intellectual content, K.D.H.; literature research, K.D.H., Y.L.; experimental studies, Y.L., K.A.A.; data acquisition, Y.L.; data analysis, K.D.H., D.T.M.; statistical analysis, D.T.M.; manuscript preparation, K.D.H. manuscript editing and review, all authors.
|
REFERENCES
|
|---|
-
Napel S, Markes MP, Rubin GD, et al. CT angiography with spiral CT and maximum intensity projection. Radiology 1992; 185:607-610.[Abstract/Free Full Text]
-
Rubin GD, Dake MD, Napel SA, McDonnell CH, Jeffrey RB. Three-dimensional spiral CT angiography of the abdomen: initial clinical experience. Radiology 1993; 186:147-152.[Abstract/Free Full Text]
-
Marks MP, Napel S, Jordan JE, Enzmann DR. Diagnosis of carotid artery disease: preliminary experience with maximum-intensity-projection spiral CT angiography. AJR Am J Roentgenol 1993; 160:1267-1271.[Abstract/Free Full Text]
-
Rubin GD, Dake MD, Napel SA, et al. Spiral CT of renal artery stenosis: comparison of three-dimensional rendering techniques. Radiology 1994; 190:181-189.[Abstract/Free Full Text]
-
Rubin GD, Paik DS, Johnston PC, Napel S. Measurement of the aorta and its branches with helical CT. Radiology 1998; 206:823-829.[Abstract/Free Full Text]
-
Berg MH, Manninen HI, Vanninen RL, Vainio PA, Soimakallio S. Assessment of renal artery stenosis with CT angiography: usefulness of multiplanar reformation, quantitative stenosis measurements, and densitometric analysis of renal parenchymal enhancement as adjuncts to MIP film reading. J Comput Assist Tomogr 1998; 22:533-540.[Medline]
-
Cumming MJ, Morrow IM. Carotid artery stenosis: a prospective comparison of CT angiography and conventional angiography. AJR Am J Roentgenol 1994; 163:517-523.[Abstract/Free Full Text]
-
Dillon EH, van Leeuwen MS, Fernanadez MA, Eikelboom BC, Maili WP. CT angiography: application to the evaluation of carotid artery stenosis. Radiology 1993; 189:211-219.[Abstract/Free Full Text]
-
Beregi JP, Elkohen M, Deklunder G, Artaud D, Coullet JM, Wattinne L. Helical CT angiography compared with arteriorgraphy in the detection of renal artery stenosis. AJR Am J Roentgenol 1996; 167:495-501.[Abstract/Free Full Text]
-
Kaatee R, Beek FJ, de Lange EE, et al. Renal artery stenosis: detection and quantification with spiral CT angiography versus optimized digital subtraction angiography. Radiology 1997; 205:121-127.[Abstract/Free Full Text]
-
Leclere X, Godefroy O, Pruvo JP, Leys D. Computed tomographic angiography for the evaluation of carotid artery stenosis. Stroke 1995; 26:1577-1581.[Abstract/Free Full Text]
-
Claves JL, Wise SW, Hopper KD, Tully D, TenHave TR, Weaver J. Evaluation of various contrast densities in the diagnosis of carotid stenosis by CT angiography. AJR Am J Roentgenol 1997; 169:569-573.[Abstract/Free Full Text]
-
Dix JE, Evans AJ, Kallmes DF, Sobel AH, Phillips CD. Accuracy and precision of CT angiography in a model of carotid artery bifurcation stenosis. AJNR Am J Neuroradiol 1997; 18:409-415.[Abstract]
-
Lev MH, Ackerman RH, Chehade RJ, et al. Accurate spiral CT angiographic measurement of carotid artery luminal diameter: a phantom study (abstr). Radiology 1996; 201(P):117.
-
Diederichs CG, Keating DP, Glatting G, Oestmann JW. Blurring of vessels in spiral CT angiography: effects of collimation width, pitch, viewing plane and windowing in maximum intensity projection. J Comput Assist Tomogr 1996; 20:965-974.[Medline]
-
Vandermeulen D, Van Hoe L, Marchal G, Suetens P, Baert AL. Assessment of blood vessel diameter measurements in spiral CT (abstr). Radiology 1994; 193(P):171.
-
Van Hoe L, Vandermeulen D, Gryspeerdt S, et al. Assessment of accuracy of renal artery stenosis grading in helical CT angiography using maximum intensity projections. Eur Radiol 1996; 6:658-664.[Medline]
-
Davros WJ, Obuchowski NA, Berman PM, Zeman RK. A phantom study: evaluation of renal artery stenosis using helical CT and 3D reconstructions. J Comput Assist Tomogr 1997; 21:156-161.[Medline]
-
Brink JA, Lim JT, Wang G, Heiken JP, Deyoe LA, Vannier MW. Technical optimization of spiral CT for depiction of renal artery stenosis: in vitro analysis. Radiology 1995; 194:157-163.[Abstract/Free Full Text]
-
Wise SW, Hopper KD, Schwartz TA, Ten Have TR, Kasales CJ. Technical factors of CT angiography studied with a carotid artery phantom. AJNR Am J Neuroradiol 1997; 18:401-408.[Abstract]
-
Wise SW, Hopper KD, Ten Have T, Schwartz TA. Measuring carotid artery stenosis using CT angiography: the dilemma of artifactual lumen eccentricity. AJR Am J Roentgenol 1998; 170:919-923.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
H.M. Silvennoinen, S. Ikonen, L. Soinne, M. Railo, and L. Valanne
CT Angiographic Analysis of Carotid Artery Stenosis: Comparison of Manual Assessment, Semiautomatic Vessel Analysis, and Digital Subtraction Angiography
AJNR Am. J. Neuroradiol.,
January 1, 2007;
28(1):
97 - 103.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
M. M. Lell, K. Anders, M. Uder, E. Klotz, H. Ditt, F. Vega-Higuera, T. Boskamp, W. A. Bautz, and B. F. Tomandl
New Techniques in CT Angiography
RadioGraphics,
October 1, 2006;
26(suppl_1):
S45 - S62.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. Suzuki, S. Furui, and T. Kaminaga
Accuracy of Automated CT Angiography Measurement of Vascular Diameter in Phantoms: Effect of Size of Display Field of View, Density of Contrast Medium, and Wall Thickness
Am. J. Roentgenol.,
June 1, 2005;
184(6):
1940 - 1944.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
M. Berg, Z. Zhang, A. Ikonen, P. Sipola, R. Kalviainen, H. Manninen, and R. Vanninen
Multi-Detector Row CT Angiography in the Assessment of Carotid Artery Disease in Symptomatic Patients: Comparison with Rotational Angiography and Digital Subtraction Angiography
AJNR Am. J. Neuroradiol.,
May 1, 2005;
26(5):
1022 - 1034.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
Y.-H. Kim, E. M. Marom, J. E. Herndon II, and H. P. McAdams
Pulmonary Vein Diameter, Cross-sectional Area, and Shape: CT Analysis
Radiology,
April 1, 2005;
235(1):
43 - 49.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. Suzuki, S. Furui, T. Kaminaga, and T. Yamauchi
Measurement of Vascular Diameter In Vitro by Automated Software for CT Angiography: Effects of Inner Diameter, Density of Contrast Medium, and Convolution Kernel
Am. J. Roentgenol.,
May 1, 2004;
182(5):
1313 - 1317.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
K. A. Addis, K. D. Hopper, T. A. Iyriboz, Y. Liu, S. W. Wise, C. J. Kasales, J. S. Blebea, and D. T. Mauger
CT Angiography: In Vitro Comparison of Five Reconstruction Methods
Am. J. Roentgenol.,
November 1, 2001;
177(5):
1171 - 1176.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
