DOI: 10.1148/rg.252045098
RadioGraphics 2005;25:425-438
© RSNA, 2005
Pitfalls in 16–Detector Row CT of the Coronary Arteries1
Tadashi Nakanishi, MD,
Yasuyo Kayashima, MD,
Rintaro Inoue, MD,
Kotaro Sumii, MD and
Yukihiko Gomyo, MD
1 From the Departments of Radiology (T.N., Y.K.) and Cardiology (R.I., K.S., Y.G.), Mazda Hospital, Mazda Motor Corporation, 2–15 Aosakiminami, Fuchu-cho, Aki-gun, Hiroshima 735-8585, Japan. Presented as an education exhibit at the 2003 RSNA Scientific Assembly. Received May 3, 2004; revision requested June 16 and received July 28; accepted July 29. All authors have no financial relationships to disclose.
Address correspondence to T.N. (e-mail: nakanishi.ta{at}mazda.co.jp).
 |
Abstract
|
|---|
Recently developed 16–detector row computed tomography (CT) has been introduced as a reliable noninvasive imaging modality for evaluating the coronary arteries. In most cases, with appropriate premedication that includes ß-blockers and nitroglycerin, ideal data sets can be acquired from which to obtain excellent-quality coronary CT angiograms, most often with multiplanar reformation, thin-slab maximum intensity projection, and volume rendering. However, various artifacts associated with data creation and reformation, postprocessing methods, and image interpretation can hamper accurate diagnosis. These artifacts can be related to pulsation (nonassessable segments, pseudostenosis) as well as rhythm disorders, respiratory issues, partial volume averaging effect, high-attenuation entities, inappropriate scan pitch, contrast material enhancement, and patient body habitus. Some artifacts have already been resolved with technical advances, whereas others represent partially inherent limitations of coronary CT angiography. Familiarity with the pitfalls of coronary angiography with 16–detector row CT, coupled with the knowledge of both the normal anatomy and anatomic variants of the coronary arteries, can almost always help radiologists avoid interpretive errors in the diagnosis of coronary artery stenosis.
© RSNA, 2005
Abbreviations: ECG = electrocardiography, LAD = left anterior descending, MIP = maximum intensity projection, MPR = multiplanar reformation, RCA = right coronary artery, VR = volume rendering
|
Introduction
|
|---|
Conventional coronary catheter angiography is the standard of reference for the assessment of coronary artery disease. Because catheterization involves admission to the hospital and discomfort for the patient, conventional angiography should be undertaken only for strict clinical indications. For these reasons, noninvasive imaging of the coronary arteries has attracted growing interest in the past few years. Contrast material–enhanced computed tomography (CT) of the coronary arteries is becoming feasible as the temporal and spatial resolution of CT improves with the availability of 16–detector row scanners (1). The fast volume coverage of retrospective electrocardiographically (ECG)–gated multisection CT allows acquisition of the entire heart volume with nearly isotropic resolution within a single breath hold for visualization of the coronary arteries.
There have been several reports on the usefulness of four–detector row CT (2–7) and 16–detector row CT (8,9) for the detection of coronary artery disease. On the other hand, the limitations of coronary CT angiography have been reported to include partial volume averaging effect, calcification, inhomogeneous enhancement, and motion artifacts (6). Coronary CT angiograms are usually interpreted from various postprocessed images and axial source images obtained with contrast-enhanced ECG-gated multisection CT. The pitfalls of coronary CT angiography are closely related to limiting factors and postprocessing methods (10). In this article, we describe our techniques for coronary multisection CT angiography and briefly discuss the requirements for coronary CT angiographic data sets. We also discuss and illustrate the spectrum of artifacts that can simulate coronary artery stenosis and lead to nonassessable segments and discuss postprocessing pitfalls. In addition, we review the normal anatomy and anatomic variants of the coronary arteries and discuss effective strategies for improving the diagnostic accuracy of coronary CT angiography.
|
Techniques for Coronary Multisection CT Angiography
|
|---|
Retrospective ECG-gated scans were obtained using a 16–detector row CT scanner (Somatom Sensation Cardiac; Siemens Medical Solutions, Forchheim, Germany) with a 0.42-second rotation time. For the cardiac protocol, the 12 inner detector rings were applied in retrospective ECG gating (standard protocol: pitch = 0.31; slow heart rate protocol for bpm <50: pitch = 0.29, where pitch is defined as the distance the table travels during one 360° rotation divided by the total collimated width of the x-ray beam). A 100-mL bolus of iopamidol (300 or 370 mg of iodine per milliliter) (Iopamiron; Nihon Schering, Osaka, Japan) followed by 40 mL of saline solution was intravenously injected with a power injector (Dualshot; Nemoto, Tokyo, Japan) at a rate of 4 mL/sec using autodetection of the contrast material at the left atrium. Optimum aortic attenuation was ensured by triggering scanning on the basis of identification of 100-HU attenuation in the ascending aorta.
A section thickness of 0.75 mm with 0.5-mm overlap, an 18-cm field of view, and an acquisition matrix of 512 x 512 were used. The resulting spatial resolution obtained with this technique was an almost isotropic voxel of 0.4 x 0.4 x 0.6 mm.
Patients with a prescan heart rate over 65 bpm were given a single oral dose of 40–60 mg of metoprolol 1 hour before the examination in the absence of contraindications. Nitroglycerin (0.3 mg) was given sublingually to dilate the coronary arteries. A temporal resolution of 210 msec was used in patients with a heart rate of 68 bpm or lower. Multiplanar reformation (MPR) was performed if a patient had a heart rate of 68 bpm or higher, with a resultant temporal resolution ranging from 105 to 210 msec depending on the heart rate. Initially, one data set was reformatted with the reformation window starting at a fixed time of 400 msec prior to the next R wave. If motion artifacts were present in the coronary arteries, image reformation was repeated with the reformation window offset 20 msec toward the beginning and end of the cardiac cycle until images without motion artifacts were obtained or until 10 data sets had been created. In the latter case, the data set with the fewest motion artifacts was used for further evaluation of the coronary arteries.
|
Requirements for Coronary CT Angiographic Data Sets
|
|---|
Accurate diagnosis with coronary CT angiography requires motion-free volume data sets with a high level of structural consistency. Consistency of data sets as well as motion-free images of the heart can be easily evaluated with MPR (Fig 1). High-quality coronary angiograms can be obtained from data sets with a high level of structural consistency (Fig 2). Various image degradation factors can cause two diagnostic pitfalls in coronary CT angiography: nonassessable segments of coronary arteries and pseudostenosis due to any cause. To minimize diagnostic errors from degradation factors, the consistency and quality of the reformatted image data sets should be examined before proceeding to image processing.
View larger version (155K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1a. High-quality MPR images obtained from coronary CT angiographic data sets with excellent structural consistency. The patient, a 41-year-old woman, had an average heart rate of 48 bpm during CT angiography. (a) Axial image obtained at the midventricular level shows a round enhanced area representing the right coronary artery (RCA). No motion artifact is seen. (b, c) Coronal (b) and sagittal (c) images show no undulating contours at the cardiac border.
|
|
View larger version (158K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1b. High-quality MPR images obtained from coronary CT angiographic data sets with excellent structural consistency. The patient, a 41-year-old woman, had an average heart rate of 48 bpm during CT angiography. (a) Axial image obtained at the midventricular level shows a round enhanced area representing the right coronary artery (RCA). No motion artifact is seen. (b, c) Coronal (b) and sagittal (c) images show no undulating contours at the cardiac border.
|
|
View larger version (159K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1c. High-quality MPR images obtained from coronary CT angiographic data sets with excellent structural consistency. The patient, a 41-year-old woman, had an average heart rate of 48 bpm during CT angiography. (a) Axial image obtained at the midventricular level shows a round enhanced area representing the right coronary artery (RCA). No motion artifact is seen. (b, c) Coronal (b) and sagittal (c) images show no undulating contours at the cardiac border.
|
|
View larger version (266K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2. Volume-rendered (VR) images obtained from coronary CT angiographic data sets with excellent structural consistency. The main coronary artery branches can be easily visualized once the main pulmonary artery, right atrium (RA), and left atrial appendage (LAA) are removed. VR findings can be compared with and used to validate findings at catheter angiography, since an arbitrary view angle can be set. RV = right ventricle.
|
|
|
Artifacts at Coronary CT Angiography
|
|---|
Various artifacts can degrade image quality at coronary CT angiography. Artifacts at CT angiography were categorized according to cause. We focused on pulsation, rhythm disorders, respiratory issues, partial volume averaging effect, high-attenuation entities, inappropriate scan pitch, contrast material enhancement, and patient body habitus as important factors hampering accurate diagnosis. Reports in which four–detector row CT with a rotation time of 500 msec was used had shown coronary CT angiography to have several technical limitations. Some issues have been resolved with 16–detector row CT with a rotation time of 420 msec, although other issues remain.
Pulsation
Artifacts attributable to cardiac pulsation are the most common and important factors degrading image quality at coronary artery imaging. Motion-free imaging during every cardiac cycle requires a temporal resolution of 50 msec (11). The strongest movement is present during contraction of the atria and ventricles in systole. Image quality should improve during the diastolic phase of the cardiac cycle: Less movement is present during the filling phase, so that the least amount of blurring due to motion artifact is to be expected. The greatest motion was seen for the RCA, whereas the least was seen for the left anterior descending (LAD) artery. The RCA and the left circumflex artery are closer to the atria than is the LAD artery (12); as a result, they are more affected by atrial contraction. Each coronary artery is most susceptible to motion artifacts during a particular part of the cardiac cycle. For these reasons, multiphase reformation and the selection of optimal reformation windows for evaluation are accepted practice for coronary multisection CT angiography (13). The duration of diastole depends on the heart rate and decreases as the heart rate decreases. The proportion of the cardiac cycle spent in systole increases as the heart rate increases and decreases as the heart rate decreases (14). For moderate and higher heart rates, the best results are obtained with two to three separate reformations for different coronary vessels. For a very low heart rate (< 50 bpm), use of one optimal reformation window may provide excellent data sets for all coronary arteries. Occasionally, consecutive marked motion artifacts can give the appearance of nonexistent vascular discontinuity or wall irregularity, even with several reformations of data from different cardiac phases (Fig 3). Coronary artery segments that move too fast compared with the acquired temporal resolution have a blurred or curvilinear appearance. Of the coronary arteries, the RCA most frequently demonstrates blurring. Nonassessable segments are attributable to extensive motion artifacts and section gaps, whereas pseudostenosis is produced by minor motion artifacts and section gaps (Fig 4).
View larger version (170K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3a. Pulsation artifacts. (a, b) Left anterior oblique (a) and anterior (b) MPR images clearly demonstrate a kymographic gap along the RCA with moderate motion artifacts at the mid-RCA. (c) Left anterior oblique thin-slab maximum-intensity-projection (MIP) image is suspicious for mild stenosis at the mid-RCA (arrow). (d) Anterior thin-slab MIP image shows definite motion artifacts (arrows), which appear to create pseudostenosis at the mid-RCA. In theory, MIP obscures a kymographic gap perpendicular to the slab plane, since the appearances of thin-slab MIP images depend on slab thickness and the orientation of the vessel of interest relative to the slab plane.
|
|
View larger version (132K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3b. Pulsation artifacts. (a, b) Left anterior oblique (a) and anterior (b) MPR images clearly demonstrate a kymographic gap along the RCA with moderate motion artifacts at the mid-RCA. (c) Left anterior oblique thin-slab maximum-intensity-projection (MIP) image is suspicious for mild stenosis at the mid-RCA (arrow). (d) Anterior thin-slab MIP image shows definite motion artifacts (arrows), which appear to create pseudostenosis at the mid-RCA. In theory, MIP obscures a kymographic gap perpendicular to the slab plane, since the appearances of thin-slab MIP images depend on slab thickness and the orientation of the vessel of interest relative to the slab plane.
|
|
View larger version (174K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3c. Pulsation artifacts. (a, b) Left anterior oblique (a) and anterior (b) MPR images clearly demonstrate a kymographic gap along the RCA with moderate motion artifacts at the mid-RCA. (c) Left anterior oblique thin-slab maximum-intensity-projection (MIP) image is suspicious for mild stenosis at the mid-RCA (arrow). (d) Anterior thin-slab MIP image shows definite motion artifacts (arrows), which appear to create pseudostenosis at the mid-RCA. In theory, MIP obscures a kymographic gap perpendicular to the slab plane, since the appearances of thin-slab MIP images depend on slab thickness and the orientation of the vessel of interest relative to the slab plane.
|
|
View larger version (147K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3d. Pulsation artifacts. (a, b) Left anterior oblique (a) and anterior (b) MPR images clearly demonstrate a kymographic gap along the RCA with moderate motion artifacts at the mid-RCA. (c) Left anterior oblique thin-slab maximum-intensity-projection (MIP) image is suspicious for mild stenosis at the mid-RCA (arrow). (d) Anterior thin-slab MIP image shows definite motion artifacts (arrows), which appear to create pseudostenosis at the mid-RCA. In theory, MIP obscures a kymographic gap perpendicular to the slab plane, since the appearances of thin-slab MIP images depend on slab thickness and the orientation of the vessel of interest relative to the slab plane.
|
|
View larger version (154K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4a. Pulsation artifacts. (a) Left anterior oblique coronary catheter angiogram demonstrates no significant stenosis at the RCA. Image reformation was performed to optimize coronary CT angiography. (b, c) Thin-slab MIP images created from data obtained in different cardiac cycles show different findings. In data sets with the reformation window starting 400 msec prior to the next R wave, apparent section gaps are observed (arrowheads in b), whereas mild stenosis seems to exist in data sets with the reformation window starting 200 msec prior to the next R wave (arrowhead in c).
|
|
View larger version (146K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4b. Pulsation artifacts. (a) Left anterior oblique coronary catheter angiogram demonstrates no significant stenosis at the RCA. Image reformation was performed to optimize coronary CT angiography. (b, c) Thin-slab MIP images created from data obtained in different cardiac cycles show different findings. In data sets with the reformation window starting 400 msec prior to the next R wave, apparent section gaps are observed (arrowheads in b), whereas mild stenosis seems to exist in data sets with the reformation window starting 200 msec prior to the next R wave (arrowhead in c).
|
|
View larger version (161K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4c. Pulsation artifacts. (a) Left anterior oblique coronary catheter angiogram demonstrates no significant stenosis at the RCA. Image reformation was performed to optimize coronary CT angiography. (b, c) Thin-slab MIP images created from data obtained in different cardiac cycles show different findings. In data sets with the reformation window starting 400 msec prior to the next R wave, apparent section gaps are observed (arrowheads in b), whereas mild stenosis seems to exist in data sets with the reformation window starting 200 msec prior to the next R wave (arrowhead in c).
|
|
Rhythm Disorders
Rhythm disorders involve apparent arrhythmia such as atrial fibrillation. In spite of recent technical advances, reliable coronary artery imaging with CT can be performed only in patients with normal sinus rhythm at the time of examination. Even in these patients, however, alteration of the heart rate during data acquisition can create data sets in a slightly different cardiac cycle, leading to a section gap and causing pseudostenosis.
Section gaps exaggerate the rapid moving segment and higher heart rate. Although a significant heart rate increase leading to a section gap cannot be determined, the association of section gaps and heart rate increase is frequently recognized (Fig 5). Nonassessable segments are frequently observed in arrhythmia, although successful coronary CT angiography can be performed even in atrial fibrillation, when the average heart rate is very low (Fig 6). Apparent multiple section gaps in imaging data sets are called banding artifacts. Banding artifacts are observed even in motion-free imaging data sets. The most frequent causes of banding artifacts are arrhythmia, no breath hold, and alterations in heart rate during acquisition. However, the occurrence of banding artifacts cannot be predicted prior to scanning in most cases. Long acquisition time is associated with increased frequency of alteration in heart rate during data acquisition. Thus, 16–detector row CT, with its fast scanning capability, reduces the prevalence and degree of alterations in heart rate during data acquisition.
View larger version (127K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5a. Artifacts due to increased heart rate in a 46-year-old woman. The patient experienced alterations in heart rate in normal sinus rhythm during scanning, which was performed shortly after the sublingual administration of nitroglycerin. The patient’s average heart rate was 51 bpm, increasing to 69 bpm in the last third of the acquisition. Coronal (a) and sagittal (b) reformatted images of the heart obtained from CT data demonstrate banding artifacts (arrowheads), which were observed only in the last third.
|
|
View larger version (152K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5b. Artifacts due to increased heart rate in a 46-year-old woman. The patient experienced alterations in heart rate in normal sinus rhythm during scanning, which was performed shortly after the sublingual administration of nitroglycerin. The patient’s average heart rate was 51 bpm, increasing to 69 bpm in the last third of the acquisition. Coronal (a) and sagittal (b) reformatted images of the heart obtained from CT data demonstrate banding artifacts (arrowheads), which were observed only in the last third.
|
|
View larger version (147K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6a. Excellent coronary angiograms of the RCA (a) and LAD artery (b) obtained in a 65-year-old woman in atrial fibrillation. The fibrillation was causing a variable R-R interval, probably because of the relatively long R-R interval.
|
|
View larger version (147K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6b. Excellent coronary angiograms of the RCA (a) and LAD artery (b) obtained in a 65-year-old woman in atrial fibrillation. The fibrillation was causing a variable R-R interval, probably because of the relatively long R-R interval.
|
|
Respiratory Issues
Breath-hold instructions and patient cooperation are essential for successful CT of the coronary arteries. Strict and repeated instructions are needed to obtain reliable coronary CT angiograms. The duration of breath holding can be reduced by using 16–detector row CT instead of four– or eight–detector row CT. It is frequently difficult to tell whether banding artifacts are caused by alterations in heart rate or by incomplete breath holding. Thus, even if a considerable amount of time is spent trying to reduce banding artifacts by searching for an appropriate cardiac window, such reduction may be impossible because the artifacts are in fact caused by incomplete breath holding (Fig 7 ).
View larger version (150K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7a. Artifacts due to incomplete breath holding. (a) Contiguous sections demonstrate almost no motion artifacts when each image is observed separately. (b, c) Coronal (b) and sagittal (c) reformatted images demonstrate banding artifacts with kymographic contours at the cardiac border. The patient had a heart rate of 58 bpm in normal sinus rhythm during scanning. Later, it was discovered that a microphone in the CT room was out of order and that breath-holding instructions had not been given to the patient.
|
|
View larger version (121K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7b. Artifacts due to incomplete breath holding. (a) Contiguous sections demonstrate almost no motion artifacts when each image is observed separately. (b, c) Coronal (b) and sagittal (c) reformatted images demonstrate banding artifacts with kymographic contours at the cardiac border. The patient had a heart rate of 58 bpm in normal sinus rhythm during scanning. Later, it was discovered that a microphone in the CT room was out of order and that breath-holding instructions had not been given to the patient.
|
|
View larger version (140K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7c. Artifacts due to incomplete breath holding. (a) Contiguous sections demonstrate almost no motion artifacts when each image is observed separately. (b, c) Coronal (b) and sagittal (c) reformatted images demonstrate banding artifacts with kymographic contours at the cardiac border. The patient had a heart rate of 58 bpm in normal sinus rhythm during scanning. Later, it was discovered that a microphone in the CT room was out of order and that breath-holding instructions had not been given to the patient.
|
|
Partial Volume Averaging Effect
Partial volume averaging effect is inherent in cross-sectional imaging modalities such as CT and magnetic resonance imaging. In the evaluation of vascular disease, the assessability of vessel diameter depends on spatial resolution along the z axis. With regard to the coronary arteries, a segment greater than 1.5 mm in diameter is considered a target for evaluation, since interventional procedures are not applied to small-diameter segments (8,9). The administration of nitroglycerin helps expand the vessel lumen and improve vessel visibility. Entities with very high attenuation around the lumina of the coronary arteries include arterial calcification, mitral annulus calcification, and intracoronary stents. Such high-attenuation entities are likely to cause blooming artifacts, leading to nonassessable segments or pseudostenosis. There are several reports on the limitations of CT angiography in the evaluation of instent stenosis (15,16). Partial volume averaging effect influences coronary artery imaging in two ways: (a) the blooming effect on the coronary lumen leads to pseudostenosis or nonassessable segments due to calcification or intracoronary stent, and (b) the plaque attenuation depends on imaging plane or section thickness. The partial volume averaging effect caused by high-attenuation entities exaggerates luminal stenosis according to the degree of attenuation (Figs 8, 9).
View larger version (165K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8a. Partial volume averaging effect. (a) Right anterior oblique catheter angiogram of the left coronary artery shows no stenosis. (b, c) MPR images depict no significant stenosis but do show a small calcification, which appears larger than it proved to be in reality. (d) Virtual endoscopic image of the coronary arteries reveals a calcified deposit at the LAD artery (arrow) without luminal stenosis.
|
|
View larger version (155K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8b. Partial volume averaging effect. (a) Right anterior oblique catheter angiogram of the left coronary artery shows no stenosis. (b, c) MPR images depict no significant stenosis but do show a small calcification, which appears larger than it proved to be in reality. (d) Virtual endoscopic image of the coronary arteries reveals a calcified deposit at the LAD artery (arrow) without luminal stenosis.
|
|
View larger version (154K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8c. Partial volume averaging effect. (a) Right anterior oblique catheter angiogram of the left coronary artery shows no stenosis. (b, c) MPR images depict no significant stenosis but do show a small calcification, which appears larger than it proved to be in reality. (d) Virtual endoscopic image of the coronary arteries reveals a calcified deposit at the LAD artery (arrow) without luminal stenosis.
|
|
View larger version (146K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8d. Partial volume averaging effect. (a) Right anterior oblique catheter angiogram of the left coronary artery shows no stenosis. (b, c) MPR images depict no significant stenosis but do show a small calcification, which appears larger than it proved to be in reality. (d) Virtual endoscopic image of the coronary arteries reveals a calcified deposit at the LAD artery (arrow) without luminal stenosis.
|
|
View larger version (175K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9a. Partial volume averaging effect. (a, b) Coronary catheter angiograms show wall irregularity of the LAD artery (arrowheads) without significant stenosis. (c, d) On thin-slab MIP (c) and curved MPR (d) images, the patency of the vessel lumen in the coronary arteries is difficult to appreciate due to diffuse and dense calcification. This blooming effect can lead to the creation of nonassessable segments or to pseudostenosis, depending on interpretation.
|
|
View larger version (173K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9b. Partial volume averaging effect. (a, b) Coronary catheter angiograms show wall irregularity of the LAD artery (arrowheads) without significant stenosis. (c, d) On thin-slab MIP (c) and curved MPR (d) images, the patency of the vessel lumen in the coronary arteries is difficult to appreciate due to diffuse and dense calcification. This blooming effect can lead to the creation of nonassessable segments or to pseudostenosis, depending on interpretation.
|
|
View larger version (148K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9c. Partial volume averaging effect. (a, b) Coronary catheter angiograms show wall irregularity of the LAD artery (arrowheads) without significant stenosis. (c, d) On thin-slab MIP (c) and curved MPR (d) images, the patency of the vessel lumen in the coronary arteries is difficult to appreciate due to diffuse and dense calcification. This blooming effect can lead to the creation of nonassessable segments or to pseudostenosis, depending on interpretation.
|
|
View larger version (124K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9d. Partial volume averaging effect. (a, b) Coronary catheter angiograms show wall irregularity of the LAD artery (arrowheads) without significant stenosis. (c, d) On thin-slab MIP (c) and curved MPR (d) images, the patency of the vessel lumen in the coronary arteries is difficult to appreciate due to diffuse and dense calcification. This blooming effect can lead to the creation of nonassessable segments or to pseudostenosis, depending on interpretation.
|
|
Streak Artifacts from High-Attenuation Entities
So-called streak artifacts are generated by high-attenuation entities, high-contrast interfaces, and cardiac motion. With retrospective ECG-gated acquisition, streak artifacts caused by high-contrast interfaces and cardiac motion can be minimized (17). Calcification and certain types of coronary artery stents with very high attenuation may produce streak artifacts, which hamper visualization of the coronary arteries. Highly concentrated contrast material in the superior vena cava can cause substantial streak artifacts that obscure the RCA. In 16–detector row CT, with which the whole heart can be imaged in about 20 seconds, the streak artifacts in the superior vena cava due to high-attenuation contrast material have been reduced with appropriate scan timing and the administration of saline solution (Fig 10). Extremely dense and extensive calcification is one of the causes of streak artifacts as well as partial volume averaging effect. The nature and extent of artifacts caused by intracoronary stents depend largely on the material of which the stent is made (Figs 11, 12). Lumen visibility varies depending on the type of stent; stents made of or coated with gold cause the most severe artifacts (15).
View larger version (132K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10a. Streak artifacts. (a) On an image obtained after the administration of saline solution, no streak artifacts are observed around the RCA because attenuation of the right atrial appendage is lowered. (b) On an image obtained in a different patient without the administration of saline solution, streak artifacts produced by high-attenuation contrast material at the right atrial appendage (arrow) affect the visibility of the proximal RCA.
|
|
View larger version (122K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10b. Streak artifacts. (a) On an image obtained after the administration of saline solution, no streak artifacts are observed around the RCA because attenuation of the right atrial appendage is lowered. (b) On an image obtained in a different patient without the administration of saline solution, streak artifacts produced by high-attenuation contrast material at the right atrial appendage (arrow) affect the visibility of the proximal RCA.
|
|
View larger version (133K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11a. (a) Coronary catheter angiogram obtained after insertion of a Radius stent (SciMED Live Systems, Maple Grove, Minn) (arrowheads) depicts a patent RCA. (b, c) Curved planar reformatted images obtained in two perpendicular planes from CT data clearly demonstrate patency of the mid-RCA with coronary stent insertion. Note that almost no streak artifacts or blooming artifacts are observed.
|
|
View larger version (113K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11b. (a) Coronary catheter angiogram obtained after insertion of a Radius stent (SciMED Live Systems, Maple Grove, Minn) (arrowheads) depicts a patent RCA. (b, c) Curved planar reformatted images obtained in two perpendicular planes from CT data clearly demonstrate patency of the mid-RCA with coronary stent insertion. Note that almost no streak artifacts or blooming artifacts are observed.
|
|
View larger version (148K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11c. (a) Coronary catheter angiogram obtained after insertion of a Radius stent (SciMED Live Systems, Maple Grove, Minn) (arrowheads) depicts a patent RCA. (b, c) Curved planar reformatted images obtained in two perpendicular planes from CT data clearly demonstrate patency of the mid-RCA with coronary stent insertion. Note that almost no streak artifacts or blooming artifacts are observed.
|
|
View larger version (137K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 12a. Streak artifacts in a patient who had undergone stent placement 5 years earlier, although the kind of stent is unknown. (a, b) Thin-slab MIP (a) and MPR (b) images show apparent streak artifacts caused by a coronary stent in the left circumflex artery (arrows in a) despite the fact that almost no motion artifacts are seen. Streak artifacts due to very high attenuation contrast material completely obscure the intracoronary lumen. (c) Even on a CT scan obtained with a very wide window setting, only metallic structures can be recognized.
|
|
View larger version (147K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 12b. Streak artifacts in a patient who had undergone stent placement 5 years earlier, although the kind of stent is unknown. (a, b) Thin-slab MIP (a) and MPR (b) images show apparent streak artifacts caused by a coronary stent in the left circumflex artery (arrows in a) despite the fact that almost no motion artifacts are seen. Streak artifacts due to very high attenuation contrast material completely obscure the intracoronary lumen. (c) Even on a CT scan obtained with a very wide window setting, only metallic structures can be recognized.
|
|
View larger version (95K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 12c. Streak artifacts in a patient who had undergone stent placement 5 years earlier, although the kind of stent is unknown. (a, b) Thin-slab MIP (a) and MPR (b) images show apparent streak artifacts caused by a coronary stent in the left circumflex artery (arrows in a) despite the fact that almost no motion artifacts are seen. Streak artifacts due to very high attenuation contrast material completely obscure the intracoronary lumen. (c) Even on a CT scan obtained with a very wide window setting, only metallic structures can be recognized.
|
|
Inappropriate Scan Pitch
To achieve full volume coverage, the image stacks reformatted in subsequent cardiac cycles must cover all positions. The pitch that can be used for image acquisition is limited by the patient’s R-R interval. Standard pitch is used to scan patients with a heart rate over 50 bpm. In patients with a long R-R interval and a heart rate below 50 bpm, scanning must be performed at a low pitch to completely cover the coronary arteries. A shortage of data occurs with inappropriate scan pitch and unpredictable bradycardia (Fig 13). Image reformation yields blurred images without motion artifacts, which is specific for artifacts due to data shortage (Fig 14). Moreover, pseudostenosis may result from a data shortage caused by inappropriate scan pitch (Fig 14 ). Because there is occasional variation in the R-R interval, low heart rate can occasionally be encountered during scanning even if ECG is performed beforehand.
View larger version (37K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13. Graph illustrates how inappropriate scan pitch and bradycardia create structural gaps (and, thus, artifacts) due to the shortage of data from the same cardiac cycle. Recon = reconstruction (reformation). At coronary artery imaging, data sets from the same cardiac cycle should be obtained from the entire heart.
|
|
View larger version (145K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 14a. Image blurring due to data shortage. A 60-year-old man underwent coronary CT angiography at standard pitch, during which his heart rate was 48 bpm. (a) Contiguous sections obtained at the level of the LAD and left circumflex arteries show blurring (arrowheads at left) in the left coronary artery as well as in the ascending aorta, pulmonary artery, right ventricle, and left atrium. Note that the images on the right show no motion artifacts. (b) Thin-slab MIP image shows pseudostenosis (arrow) and blurring (arrowheads) of the RCA. (c) Coronary catheter angiogram helps confirm the findings in b.
|
|
View larger version (160K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 14b. Image blurring due to data shortage. A 60-year-old man underwent coronary CT angiography at standard pitch, during which his heart rate was 48 bpm. (a) Contiguous sections obtained at the level of the LAD and left circumflex arteries show blurring (arrowheads at left) in the left coronary artery as well as in the ascending aorta, pulmonary artery, right ventricle, and left atrium. Note that the images on the right show no motion artifacts. (b) Thin-slab MIP image shows pseudostenosis (arrow) and blurring (arrowheads) of the RCA. (c) Coronary catheter angiogram helps confirm the findings in b.
|
|
View larger version (153K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 14c. Image blurring due to data shortage. A 60-year-old man underwent coronary CT angiography at standard pitch, during which his heart rate was 48 bpm. (a) Contiguous sections obtained at the level of the LAD and left circumflex arteries show blurring (arrowheads at left) in the left coronary artery as well as in the ascending aorta, pulmonary artery, right ventricle, and left atrium. Note that the images on the right show no motion artifacts. (b) Thin-slab MIP image shows pseudostenosis (arrow) and blurring (arrowheads) of the RCA. (c) Coronary catheter angiogram helps confirm the findings in b.
|
|
Contrast Material Enhancement and Patient Body Habitus
Autodetection of contrast material arrival and an adequate amount and rate of injection produce excellent coronary artery enhancement with minimal enhancement of the coronary vein. Superimposed structures like the coronary vein are considered to be an infrequent cause of artifacts at 16–detector row CT.
Image quality in very obese patients is poor due to a low signal-to-noise ratio. We reformatted data sets with a section thickness of 1 mm instead of 0.75 mm in patients with a low signal-to-noise ratio. To maintain a balance, it is necessary to increase the radiation dose or the section thickness. The advantages and disadvantages of both modifications should be taken into consideration.
|
Postprocessing Pitfalls
|
|---|
VR is one of the most popular postprocessing methods for the three-dimensional reformation of CT angiographic data. For accurate measurement of the vessel diameter, the optimal setting should be strictly applied to data sets for coronary artery imaging, since the vessel caliber depends on the window width and level settings at VR (Fig 15) (10). Visualization of the coronary arteries with MIP requires segmentation of the atria and ventricles. Because segmentation requires time, thin-slab MIP is feasible in this setting, although high-attenuation contrast material prevents the evaluation of the inner lumen. Postprocessed images such as VR and MIP images do not provide sufficient information about the inner lumina of the coronary arteries, which can be concealed by high-attenuation entities such as calcification and intracoronary stents. MPR and curved planar reformation can demonstrate portions of the inner lumen adjacent to or covered by high-attenuation entities. Because these techniques depend on the cutting line, using the automatic determination of centerline is helpful. Interactive viewing of reformatted images is needed to avoid interpretation errors. Virtual endoscopy of the coronary arteries allows evaluation of the lumina from the inside and may contribute to visualization of the arteries (18). However, virtual endoscopy cannot provide quantitative information because it is a perspective rendering method.
View larger version (91K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 15. VR images obtained with different window width and level settings show the LAD artery and a diagonal branch (arrow). Note that the change in the caliber of the diagonal branch is greater than that in the caliber of the LAD artery, indicating that smaller-caliber vessels tend to be more affected by the use of inappropriate settings.
|
|
|
Anatomy of the Coronary Arteries
|
|---|
Knowledge of the anatomy of the coronary arteries is required for correct interpretation of coronary CT angiographic findings. The left main coronary artery arises from the left sinus of Valsalva and divides into two major branches, the LAD artery and the circumflex artery. In about 10% of cases, the circumflex artery continues as the posterior descending artery after reaching the crux of the heart. This anatomic variant of the coronary artery system is referred to as left dominant. The RCA usually branches from the posterior descending artery. Myocardial bridges manifest as a portion of a coronary artery surrounded by left-sided myocardium and may simulate stenosis of the proximal LAD artery when images are reformatted during systole.
In rare instances, a coronary artery may have an anomalous origin and course (Fig 16). Such an anomaly must be accurately evaluated, since it can range from benign to life threatening (19).
View larger version (156K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 16a. Anatomic variant of the RCA in a 79-year-old woman who complained of chest pain. (a, b) Coronary CT angiograms clearly show the origin of the RCA from the left coronary sinus by simultaneously demonstrating the LAD artery and the RCA. (c, d) Images reveal that the RCA (arrows) is compressed between the ascending aorta and the right ventricle.
|
|
View larger version (170K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 16b. Anatomic variant of the RCA in a 79-year-old woman who complained of chest pain. (a, b) Coronary CT angiograms clearly show the origin of the RCA from the left coronary sinus by simultaneously demonstrating the LAD artery and the RCA. (c, d) Images reveal that the RCA (arrows) is compressed between the ascending aorta and the right ventricle.
|
|
View larger version (164K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 16c. Anatomic variant of the RCA in a 79-year-old woman who complained of chest pain. (a, b) Coronary CT angiograms clearly show the origin of the RCA from the left coronary sinus by simultaneously demonstrating the LAD artery and the RCA. (c, d) Images reveal that the RCA (arrows) is compressed between the ascending aorta and the right ventricle.
|
|
View larger version (126K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 16d. Anatomic variant of the RCA in a 79-year-old woman who complained of chest pain. (a, b) Coronary CT angiograms clearly show the origin of the RCA from the left coronary sinus by simultaneously demonstrating the LAD artery and the RCA. (c, d) Images reveal that the RCA (arrows) is compressed between the ascending aorta and the right ventricle.
|
|
|
Improving the Diagnostic Accuracy of Coronary CT Angiography
|
|---|
Coronary CT angiography is currently the imaging technique of choice for evaluating patients with suspected coronary artery disease. However, a variety of pitfalls can be encountered that mimic coronary artery stenosis. Artifacts at CT angiography can be categorized according to several criteria. Images that are reformatted from data sets with excellent structural consistency and that demonstrate no motion artifacts are essential for successful coronary CT angiography. Conversely, images that are reformatted from data sets with poor structural consistency and that demonstrate motion artifacts may lead to inconclusive evaluation or misdiagnosis. In view of the variations in coronary artery motion among patients and even within individual patients, much of the research concerning the optimization of image quality at coronary multi–detector row CT angiography has focused on the reformation window settings and the effects of heart rate. Currently, the heart rate control policy plays an important role in optimizing image quality. In addition, heavy calcification, motion artifacts, and a vessel lumen diameter less than 1.5 mm have been identified as the most important limiting factors, which still need to be addressed even when using a 16-section CT scanner. Still, the number of nonassessable vessels seen at coronary CT angiography is lower with 16–detector row CT than with four–detector row CT (6,9).
|
Conclusions
|
|---|
Coronary CT angiography is usually performed with several postprocessed images and source images. It is necessary to understand the influence of various degradation factors on each postprocessing method used in coronary CT angiography. The advantages and disadvantages of each postprocessing method should be clarified, and optimal work flow for coronary artery imaging should be determined (Fig 17). Images should be checked comprehensively during data creation, postprocessing, and interpretation. Pitfalls in interpretation are closely associated with the characteristics of each postprocessed image. Familiarity with these common pitfalls, coupled with the knowledge of both the normal anatomy and anatomic variant of the coronary arteries, can almost always help radiologists avoid interpretive errors in the diagnosis of coronary artery stenosis.
Top
Abstract
Introduction
