| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Beth Israel Deaconess Medical Center, 330 Brookline
Ave., Boston, MA 02215.
2 Department of Radiology, A.H.E.P.A. University Hospital, St. Kyriakidis,
Thessaloniki 54636, Greece.
Received December 10, 2004;
accepted after revision March 22, 2005.
Address correspondence to V. D. Raptopoulos.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. In four groups of 20 patients with chest pain each, four MDCT protocols were used based on monitoring vascular attenuation: pulmonary embolism (150 H at pulmonary artery), aortic dissection (200 H at aortic arch), chest pain (200 H at pulmonary artery), and chest pain with ECG gating (150 H at pulmonary artery). Vascular enhancement was assessed by attenuation measurements taken from locations in the pulmonary artery (n = 3) and thoracic aorta (n = 4). The appearance of the coronary artery in regard to opacification and motion was assessed on a scale of 1 to 5 (best).
RESULTS. The mean pulmonary artery and aorta attenuation (372 H and 352 H, respectively) was significantly higher (p < 0.005, Student's t test) and the number of vessel attenuation points measuring less than 200 H (1/140) was significantly smaller (p < 0.001, chi-square test) in the chest pain compared with the dissection (318 H, 310 H; 16/140), gated chest pain (304 H, 286 H; 17/14), and pulmonary embolism (302 H, 220 H; 28/140) groups. The median coronary artery visualization score was 4; the proximal regions received a significantly (p < 0.005, Mann-Whitney test) higher grade compared with the middle and distal regions (medians, 5, 4, and 2, respectively). Artifacts were noted on the gated scans.
CONCLUSION. The chest pain protocol can be used to assess both the pulmonary arteries and the thoracic aorta, whereas the ECG-gating protocol appears to be a promising adjunct for a comprehensive single chest pain protocol.
Keywords: aorta • aortic dissection • cardiovascular imaging • coronary artery disease • CT coronary arteriography • emergency radiology • MDCT angiography technique • pulmonary embolism
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
One of the major advantages of pulmonary CTA is its ability to detect other, nonvascular, causes that could explain the symptom of acute chest pain. Such causes may be related to the pulmonary parenchyma, pleura, or chest wall [4]. Acute aortic dissection is another important vascular condition that can cause acute chest pain and for which thoracic CTA is highly accurate [6, 7]. Coronary artery disease is a major cause of acute chest pain. The incidence of coronary calcification, an indicator of coronary atherosclerosis [8], has been shown to be increased in patients undergoing pulmonary CTA [5]. It is possible that patients undergoing pulmonary CTA may also be at high risk for coronary artery disease, which may account for their pain. From the clinician's perspective, differentiating the cause of chest pain resulting from PE, aortic dissection, or an acute coronary syndrome is not always clear-cut. With the advent of MDCT, ECG-gated CTA is emerging as a viable means of imaging the coronary arteries [9, 10]. ECG gating has also been shown to reduce artifacts on images of the thoracic aorta [11] and improve visualization of the pulmonary vessels [12–14] and pulmonary parenchyma [13, 14].
Different tailored CTA protocols have been used for PE, aortic dissection, and coronary artery imaging based on experience with helical CT and 4-MDCT. These scanners are relatively slow, and for a successful study there is a delicate balance between scanning area, breath-hold, slice thickness, amount of IV contrast material, injection-to-scan delay, and injection rate [6, 9, 15]. However, with the newer 8- and 16-MDCT scanners, the chest can be scanned much faster—in 5–7 sec—using 1- or 1.25-mm collimation while the value of even thinner collimation (0.5 or 0.65 mm available with the 16-MDCT scanners) for chest imaging has not been shown. Still, a delicate balance remains between scanning area, breath-hold, slice thickness, amount of IV contrast material, injection-to-scan delay, and injection rates.
Although these parameters are presently evaluated, the speed of high-detector-row MDCT has expanded the choices of scanning in different phases of contrast vascular enhancement to the point at which a single protocol could evaluate both the pulmonary arteries and the thoracic aorta. It would be ideal if the coronary arteries could be also evaluated at the same time (Raptopoulos V et al., presented at the 2003 meeting of the Radiological Society of North America and the 2004 meeting of the American Roentgen Ray Society) (Fig. 1). This would allow evaluation of all chest vessels during a single examination. With retrospective ECG gating, the chest can be scanned in 20–28 sec using 1- or 1.25-mm collimation, which is about the same breath-hold time helical scanners require to scan the chest with 2.5- or 3-mm collimation.
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Scanning Techniques
Scans were obtained on an 8-MDCT scanner (LightSpeed Ultra, GE Healthcare)
or a 16-MDCT scanner (Aquilion, Toshiba America Medical Systems). Data
acquisition collimation was 1.25 mm with the 8-MDCT scanner and 1 mm with the
16-MDCT scanner. Of the 60 nongated scans, 43 (72%) were obtained with the
8-MDCT scanner and 17 with the 16-MDCT scanner. Of the 20 gated scans, 12
(60%) were obtained with the 8-MDCT scanner and eight with the 16-MDCT
scanner. Contiguous 5-mm-thick axial and coronal images were reconstructed and
stored on PACS. Although two different scanners were used, we kept the
scanning parameters as far as collimation and table speed quite similar to
minimize confusion in the analysis of the four groups (Tables
1 and
2). With the 16-MDCT scanner,
the data acquisition was faster than with the 8-MDCT scanner. In the groups
scanned using the nongated technique, the difference between 4 and 7 sec,
respectively, is not substantial enough to produce significantly different
results. In the group scanned using the gated technique, the difference
between 17 and 25 sec is considerable, but the caudal–cranial direction
compensates for possible involuntary respiratory motion at the end of the
scanning; motion is less in the upper lobes than in the lower lobes. Data from
the gated protocol were sent to image processing workstations for multiplanar
interactive viewing. These included an ADW 4.0 (Advanced Development
Workstation, GE Healthcare) and a Vitrea 2 (Vital Images).
|
|
Nonionic IV contrast material (ioversol [Optiray 350, Mallinckrodt]) was injected at a rate of 3.5–5 mL/sec with a single-syringe power injector (EnVision CT Injector, Medrad). For the gated protocol, 150 mL was delivered, followed by 30 mL of normal saline, whereas 100 mL was used for the other three protocols. Scanning was triggered after temporal monitoring of the attenuation of the pulmonary artery or the aorta during IV contrast injection over a single level. After a 10-sec initial delay, the monitoring scanning was done either continuously with the 16-MDCT scanner (Sure-scan, Toshiba Medical Systems) or periodically, every 2.5 sec, with the 8-MDCT scanner (Smart-prepare, GE Healthcare). There was a 4-sec inherent delay from triggering to the start of scanning (Table 1).
PE group—Attenuation monitoring to trigger scanning was done at the level of the main pulmonary artery, and scanning was triggered when the enhancement value reached 100 H on the 8-MDCT scanner or the attenuation value reached 150 H on the 16-MDCT scanner. IV contrast material (100 mL) was administered at a rate of 4–5 mL/sec. Although a smaller amount of contrast material (as low as 60–80 mL) can be used for PE detection, this was not the routine in our institution at the time of the study. Scanning was performed in the caudal-to-cranial direction. The goal was to opacify the pulmonary arteries without contamination from the pulmonary veins or aorta.
Aortic dissection group—Attenuation monitoring to trigger scanning was done at the level of the aortic arch, and scanning was triggered when the enhancement value reached 100 H on the 8-MDCT scanner or the attenuation value reached 150 H on the 16-MDCT scanner. IV contrast material (100 mL) was administered at a rate of 4–5 mL/sec. Scanning was performed in the cranial-to-caudal direction from the lower neck through the chest and continued to the level of the abdominal aortic bifurcation. The goals were to opacify the thoracic aorta and continue to opacify the abdominal aorta.
Chest pain group scanned using nongated protocol—Attenuation monitoring to trigger scanning was performed at the level of the main pulmonary artery, and scanning was triggered when the enhancement value reached 150 H on the 8-MDCT scanner or the attenuation value reached 200 H on the 16-MDCT scanner. IV contrast material (100 mL) was administered at a rate of 4–5 mL/sec. Scanning was performed in the caudal-to-cranial direction and extended from the costophrenic angles to the lower neck. This protocol differs from the PE protocol by its longer anatomic coverage and longer injection-to-scan delay (higher scan-trigger threshold). The goal was to opacify both the aorta, including the proximal brachiocephalic arteries, and the pulmonary arteries.
Chest pain group scanned using gated protocol—Attenuation monitoring to trigger scanning was performed at the level of the main pulmonary artery, and scanning was triggered when the enhancement value reached 130 H on the 8-MDCT scanner or the attenuation value reached 180 H on the 16-MDCT scanner. IV contrast material (150 mL) was administered at a rate of 3.5–5 mL/sec. A 30-mL normal saline push was given at termination of the IV contrast injection. Scanning was performed in the caudal-to-cranial direction. The goal was to opacify the pulmonary arteries, the coronary arteries, and the aorta. Because all patients had acute symptoms, no medications solely for the study were administered. In particular, study participants did not receive nitroglycerin or ß-blockers. Patients with pulse rates greater than 80 bpm were excluded because the usefulness of coronary CTA has been shown to decrease as the patient's pulse rate increases [9].
Data Acquisition
The nongated scans were obtained on the 8-MDCT scanner with a 1.25 mm
x 8 collimation and table speed of 13.5 mm per 0.5-sec rotation or 27
mm/sec, allowing the scanning of an 18-cm chest span in less than 7 sec
(Table 2). With the 16-MDCT
scanner, we used a 1 mm x 16 collimation and table speed of 21.6 mm per
0.5-sec rotation or 43.2 mm/sec, allowing the scanning of an 18-cm chest span
in a little more than 4 sec.
The gated scans were obtained on the 8-MDCT scanner with a 1.25 mm x 8 collimation and table speed of 3.51 mm per 0.5-sec rotation or 7.2 mm/sec, allowing the scanning of an 18-cm chest span in 25 sec. If a longer span was needed, the patients were allowed to let their breath out slowly. With the 16-MDCT scanner, we used a 1 mm x 16 collimation and table speed of 4.3 mm per 0.4-sec rotation or 10.75 mm/sec, allowing the scanning of an 18-cm chest span in 17 sec.
Study Display
Axial and coronal images of the chest were reconstructed using a
soft-tissue algorithm for the evaluation of the pulmonary arteries, aorta, and
lung parenchyma at 2.5-mm (8-MDCT scanner) or 3-mm (16-MDCT scanner) slice
thickness and were sent to PACS for interpretation and storage. Once the data
were in the PACS, the images were interpreted using appropriate window
settings for soft tissues, vessels, and lung. Retrospective gating was applied
to the heart images. Cardiac data were reconstructed at 70% of the R-R
interval. In cases of considerable motion artifacts, data from 40%, 60%, or
80% of the R-R intervals were retrospectively reconstructed. This was done in
five of 20 gated studies (Figs.
2A and
2B). Coronary vessels were
viewed on a workstation, and selected images were saved. Images of the
coronary arteries were obtained using interactive double oblique orthogonal
views (Figs. 3A and
3B).
Maximum-intensity-projection (MIP) slabs of 4- to 7-mm thickness were recorded
for detailed vessel evaluation, and thicker volume-rendering slabs
(10–20 mm) were obtained for broader views of vessel orientation and
anatomic relationships.
|
|
|
|
Vessel Assessment
In the gated and nongated chest pain groups, two radiologists assessed the
pulmonary and coronary arteries. The coronary arteries were divided into three
regions: proximal, middle, and distal. The proximal region included the
origins of the right coronary artery (RCA) and left main coronary artery and
the horizontal portions of the left anterior descending (LAD) and left
circumflex arteries (segments 1, 5, 6, 9). The middle region included the
descending portion of these vessels (segments 2, 7, 10, 11), and the distal
region included their apical and basal distribution and the posterior
descending artery (segments 3, 4, 8, 12). These structures were assessed for
motion artifacts and vessel attenuation. The appearance of the coronary artery
segments was graded using the following 5-point scale: 1, not seen, obscured;
2, poor, probably obscured; 3, moderate, some artifacts; 4, adequate, few
artifacts; and 5, superior, no artifacts.
The adequacy of enhancement of the pulmonary artery and aorta was determined when attenuation was greater than 200 H. Homogeneity of the contrast column in the pulmonary artery tree was assessed by measuring attenuation at three levels of the pulmonary artery: first, the main branch; second, the subsegmental branch to the right lower lobe between the level of the inferior pulmonary vein and the diaphragm; and, third, the subsegmental branch to the right upper lobe at the level of the superior portion of the aortic arch [15]. For the thoracic aorta, attenuation measurements were performed in four locations: the ascending and descending thoracic aorta at the level of the main pulmonary artery; the aortic arch; and the level of the diaphragm. Variability of the contrast column in the aorta and the pulmonary artery of each patient was assessed as the greatest difference in attenuation between the three pulmonary artery and four aorta measurements. The smaller the difference, the more homogeneous the contrast column in the pulmonary artery or aorta was considered.
Statistical Analysis
The attenuation and homogeneity of the contrast column in the pulmonary
arteries and aorta were compared in all groups. In addition, comparison of the
quality of the appearance of the coronary artery in gated and nongated chest
pain protocols was made. Comparison of coronary visualization in the proximal,
middle, and distal coronary artery was made for the gated group. The analysis
of variance was used to compare ordinal data of the various groups, such as
differences in attenuation measurements, and the Student's t test was
used to compare differences of the same data between two groups. For ordinal
data such as grading of coronary artery segments, the Mann-Whitney rank sum
test was used to compare differences between two groups. For nominal data,
such as the number of attenuation measurement points with values smaller or
larger than 250 H in the pulmonary artery or thoracic aorta, the chi-square
test was used. A p value of less than 0.05 was considered
statistically significant.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Vessel Enhancement
All examinations were satisfactory with a mean attenuation in each of the
four groups greater than 250 H in all the critical vessels
(Fig. 4). The mean attenuation
of both the pulmonary artery (372 ± 98 H) and the thoracic aorta (352
± 84 H) was statistically significantly (p < 0.001) higher
in the chest pain protocol group than in the other three groups
(Table 3). In this group,
vessel attenuation was consistently higher throughout the pulmonary arteries
and the aorta, and we observed a significantly smaller number of measurements
with attenuation values < 250 H and < 200 H
(Table 4). Of the 140 combined
pulmonary artery and thoracic aorta measurement points in each group, the
attenuation was < 200 H in one patient in the chest pain group (0.7%; 95%
CI, 0.002–0.04) compared with 16 in the dissection (11%; 95% CI,
0.07–0.2; p < 0.001), 28 in the PE (20%; 95% CI,
0.14–0.28), and 17 in the ECG-gated chest pain (12%; 95% CI,
0.07–0.19) groups.
|
|
Considering the mode of scanning (caudal–cranial), the timing of scanning, and the injection rates, scanning was early in the PE group, capturing mainly the pulmonary arteries. Scanning in the chest pain group was done during a brief pulmonary artery–aorta phase. This phase was prolonged by reducing the IV contrast injection rate in the chest pain group undergoing the gated protocol. In the dissection protocol, scanning was performed a little later than in the other groups and resulted in good consistent aortic contrast concentration that was high enough at the distal thoracic aorta to allow continued scanning in the upper abdominal aorta (Fig. 4).
The mean pulmonary artery attenuation in the PE protocol was 301 ± 69 H, and the mean thoracic aorta attenuation in the dissection protocol was 310 ± 84 H. Respectively, these were not significantly different from the mean attenuation of the pulmonary arteries (304 ± 120 H) or thoracic aorta (287 ± 96 H) in the gated group (Table 3). However, the contrast column homogeneity (least variation in attenuation at various points) was best in the pulmonary arteries (p < 0.004) of the PE group (10 ± 69 H) and in the aorta (p < 0.001) of the dissection group (11 ± 62 H) and was significantly different from the other groups: p < 0.04 and p < 0.001, respectively (Table 5).
|
Coronary arteries—In general, the coronary arteries in the gated group were fairly well visualized (Table 6). On a scale of 1–5, the mean ± SD coronary artery score was 3.4 ± 1.5 and a median of 4 (quartiles 2 and 5). The proximal coronary arteries received a significantly (p < 0.005) higher grade than the middle region (Figs. 2A, 2B, 3A, 3B, and 5). The latter also received a significantly higher grade than the distal coronary artery regions (Figs. 3B and 5). The respective mean ± SD and median values for the proximal regions were 4.3 ± 1 with a median of 5 (quartiles 4 and 5); for the middle regions, 3.3 ± 1.4 with a median of 4 (quartiles 2 and 4); and for the distal coronary regions, 2.5 ± 1.4 with a median of 2 (quartiles 1 and 4). The mid and distal regions of the LAD received significantly higher scores than the middle (p < 0.02) and distal (p < 0.03) regions of the RCA. The respective mean ± SD and median values for the mid LAD regions are 3.9 ± 1.2 and 4 compared with 2.7 ± 1.6 and 3 for the mid RCA regions; for the distal LAD regions, 3 ± 1.3 and 3 compared with 2 ± 1.4 and 1 for the distal RCA regions.
|
|
In general, the coronary arteries in the nongated groups were visualized poorly. Compared with the nongated chest pain protocol, retrospective gating significantly (p < 0.001) improved visualization of the coronary arteries. On a scale of 1–5, the mean ± SD coronary artery score on the nongated chest pain protocol was 1.7 ± 1.1 with a median of 2 (quartiles 1 and 2). In this group, the proximal regions received significantly (p < 0.001) higher scores than the mid and distal regions. The respective mean ± SD and median (and quartiles) values for the proximal vessel regions was 2.4 ± 1.3 and 2 (quartiles 1 and 2); mid vessel regions, 1.4 ± 0.7 and 1 (quartiles 1 and 2); and distal coronary artery regions, 1.2 ± 0.6 and 1 (quartiles 1 and 1).
Pulmonary arteries, aorta, and pulmonary parenchyma on ECG-gated and nongated scans—In 14 of the 20 patients who had an ECG-gated study, stepladder cardiac pulsation artifacts [9, 17, 20] of the distal lower lobe pulmonary arteries were noted on coronal multiplanar reconstruction images, ranging from a slightly serrated appearance to complete fragmentation of the vessels (Figs. 6A, 6B, 6C, and 6D). This finding appeared to be related to the reconstruction timing in the R-R interval and the pulse rate (Figs. 7A and 7B). It was not observed in the non–retrospective-gated series. It was much less pronounced in the central and upper lobe pulmonary arteries and the aorta. The ascending aorta was free of counterpulsation artifact [16, 17] in 19 of the 20 patients (Fig. 6B). The axial images were acceptable on the gated protocol scans. All nongated protocols yielded similar high-quality axial and multiplanar images.
|
|
|
|
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
It is not rare for a patient with chest pain to be referred for evaluation for both PE and aortic dissection. Traditionally, evaluation for these diagnoses has required scanning with different protocols. This has been common practice because of the relatively slow helical CT or 4-MDCT scanners. In this study, we have shown that, with the use of faster scanners, it is now possible to combine the PE and thoracic aortic dissection protocols to a single chest pain study to evaluate both conditions along with nonvascular abnormalities of the chest. Although we limited the number of variables investigated, further refinements of the technique could be made by studying the effect of detector collimation [23], the amount and iodine concentration of the contrast material [24], various injection rates, and bolus threshold detection sites.
In this study, the chest pain protocol was performed in a caudal–cranial direction, as is routinely done with the PE protocol. However, in patients with suspected aortic dissection, scanning was performed in the cranial–caudal direction and then continued to the abdomen. If the new chest pain protocol is to replace the PE and aortic dissection protocols, the problem of scanning the abdomen needs to be solved. In our study, we showed that the contrast column in the chest pain protocol was homogeneous in both the pulmonary arteries and aorta, showing little attenuation differences in measurements made in the lower and upper chest (Fig. 4 and Table 5). This would permit scanning in whichever direction seems appropriate. Presently in our practice, we scan patients with tachypnea in the caudal–cranial direction to reduce respiratory artifacts in the lower lobes, which may be severe. For patients without tachypnea who can hold their breath, we scan in a cranial–caudal direction. If the possibility of aortic dissection is raised, we follow that scan with one of the abdomen.
Coronary artery disease is another cause of acute chest pain. A subset of patients presents with atypical symptoms that may mimic PE or aortic dissection. Although coronary artery disease is a major cause of acute chest pain, these patients are not traditionally referred for chest CTA. Frequently, however, patients with coronary artery disease may present with atypical symptoms and thus may be referred to undergo CTA for evaluation of suspected PE or dissection. From the clinician's perspective, differentiating patients whose chest pain is caused by PE, aortic dissection, or an acute coronary syndrome is not always clear-cut. Prologo et al. [3] found that over the years there was increased utilization of pulmonary CTA but a decrease in the number of both PEs and ancillary findings detected, suggesting that physicians are using pulmonary CTA to screen patients with suspected cardiovascular disease. Kiryu et al. [5] showed that the prevalence of coronary artery calcification in patients undergoing non–ECG-gated pulmonary CTA was increased (was increased or was higher compared to age- and sex-matched controls). Compared with age- and sex-matched controls, those patients were 2.9 times more likely to have coronary calcification. Given that coronary artery calcification indicates coronary atherosclerosis [8], patients undergoing pulmonary CTA may be at high risk for coronary artery disease, which may account for their pain. Although the exact number of our 80 patients whose pain was caused by coronary disease is unknown, 19 (24%) required subsequent coronary artery disease evaluation, three had subsequent coronary intervention, and another three had undergone a previous coronary intervention. Of our 20 patients who underwent imaging with retrospective ECG gating, atherosclerotic plaques were detected in six. Clearly, new advances in MDCT allow the potential for expanding the use of coronary CTA [25, 26].
In this study, we have shown that a chest pain protocol, with the use of an 8- or 16-MDCT scanner, yields vascular studies that are superior to those obtained with a specific PE or aortic dissection protocol. Compared with the other three protocols, the chest pain regimen produces a higher concentration of IV contrast material in both the pulmonary arteries and the thoracic aorta (Fig. 4). This protocol, as used in this study, requires a slightly larger amount of IV contrast material than is currently used for PE (100 mL, compared with 80 mL) because of the longer injection-to-scan delay and longer body coverage, but it allows a comprehensive simple protocol. Although the diagnostic yield of higher vascular attenuation has not been studied with this protocol, we observed significantly larger numbers of measurement points with an attenuation of greater than 200 and 250 H (Table 4). In our experience, the range between 200 and 250 H is adequate and 220–350 H is optimal. In addition, retrospective gating appears to be a promising adjunct that, with additional refinements, may be used as a generic but comprehensive chest pain protocol that also includes coronary evaluation.
The preliminary results with the use of the gated chest pain protocol showed adequate opacification of the pulmonary arteries and aorta (Fig. 4 and Table 4). In addition, there was good visualization of the coronary arteries, especially the proximal portions, as shown by others [27]. This ability has an important potential application in the evaluation of type A thoracic aortic dissection, a condition for which determining whether the dissection extends into the coronary arteries is important. Both calcified and noncalcified coronary artery plaques were noted. In contrast, on nongated scans, visualization of proximal and distal coronary arteries was sporadic and was significantly inferior to that provided by images obtained using the gated technique.
The gated protocol has a number of limitations. Retrospective gating increases radiation exposure. Data from 4-MDCT show a 3- to 4-fold increase compared with exposure from routine nongated chest CTA [21, 28, 29]. This can be reduced by as much as 48% using ECG-controlled tube current modulation [30] or prospective gating with a nonhelical scanning mode [13, 14]. Although 8- and 16-MDCT scanners allow a more efficient beam utilization compared with a 4-MDCT scanner, the gated techniques still deliver a relatively higher dose, by a factor of about 3, because of the low pitch used (0.25–0.35).
There are additional limitations that newer technology may resolve. Presently, the scanning acquisition time is relatively long, requiring long breath-holds in patients who may have difficulty with breath-holding because of pain or underlying vascular or pulmonary disease. In addition, the technique cannot be applied in patients with tachycardia or an irregular pulse rate, findings that are relatively common in patients with acute chest pain. Beta-blockers or nitroglycerine may improve the quality of coronary imaging by decreasing the pulse rate and dilating the vessels, which may improve temporal and spatial resolution, respectively, but administration of these agents may not be practical for routine use in patients with acute chest pain for whom little medical history is available. In this study, we used 1- and 1.25-mm detector thickness.
Newer 64-MDCT scanners are faster and have thinner detectors, which could provide the opportunity to examine larger areas with a higher resolution in a fraction of the time required for the 8- and 16-MDCT systems. In addition, a number of software and hardware improvements could decrease "stepladder" or "band" artifact while allowing scanning of patients with pulse rates of more than 80 bpm.
Other investigators have shown cardiac gating to produce high-quality images of the pulmonary arteries, lung parenchyma, and aorta with decreased artifacts [11–14]. A pseudodissection spiral artifact of the ascending thoracic aorta has been previously described, and segmental reconstruction has been proposed to correct this [16, 17]. Retrospective gating eliminated this artifact on the axial images in our series, similar to the experience of Roos et al. [11]. In our study, we observed few motion artifacts on the axial images. However, on the multiplanar images, we observed bandlike or stepladder artifacts of varying severity in a considerable number of scans (Figs. 6A, 6B, 6C, 6D, 7A, and 7B). We attribute these artifacts to segmental data collection during continuous table motion [11, 25, 30, 31]. These artifacts have been shown to be more pronounced with increasing heart rate and are also influenced by incorrect pitch or time reconstruction point in the R-R interval [9, 20, 21]. Although this has been worked up for the coronary arteries, little attention has been paid to the pulmonary vessels (Figs. 2A, 2B, 7A, and 7B). Marten et al. [12] showed good pulmonary artery imaging using a 70% R-R interval. For decreasing cardiac pulsation motion artifacts, the best results have been reported from studies in which prospective gating with nonhelical scanning was used [13, 14]. Even in those studies, however, the clinical usefulness of the improved resolution was questionable.
There are two additional limitations associated with the gated protocol. Finding an atherosclerotic plaque in a patient with chest pain does not necessarily mean that the symptoms are due to coronary disease. From the clinical perspective, knowing that a patient with chest pain has coronary artery disease may influence his or her further evaluation and disposition. With the present methods, study execution and interpretation are time-consuming, especially if medications are given.
Despite its limitations, our gated chest pain protocol produced high-quality images of the thorax, the lungs, the pulmonary artery, and the aorta along with adequate coronary artery visualization, and in selected patients, it may be useful for the evaluation of acute atypical chest pain. However, with the present technology and until the usefulness of coronary CTA is shown, the technique is limited by long acquisition time, time-consuming image processing, high radiation dose, and pulmonary parenchyma and peripheral pulmonary artery artifacts. It should therefore be used selectively. In particular, the long acquisition time makes the technique prohibitive in patients with tachypnea, while medications may be needed in patients with tachycardia, both common symptoms accompanying chest pain. The new 64-MDCT scanners can scan the heart with ECG gating (0.5-mm collimation) followed by scanning the chest (1- to 2-mm collimation) without ECG gating in approximately 20–25 sec, including a 4-sec breath-hold between series, using 120–150 mL of contrast material injected at a rate of 5 mL/sec to yield a biphasic study with dedicated gated coronary and nongated combined vascular and pulmonary series. The feasibility of this protocol needs to be investigated further.
This is a feasibility study that suffers from a number of biases. Although each of our groups consisted of consecutive patients, selection bias was unavoidable because the patients were selected in each group according to clinical symptoms that may have been misleading. In addition, the EGG-gated chest pain group is further skewed from the nongated group secondary to age and heart rate criteria and scanning time-of-the-day availability. These limitations were unavoidable in this technique-development study, but the results are encouraging to justify additional studies with stricter protocols. Furthermore, there is a verification bias because the results of the test have not been compared with a reference test and a diagnostic review bias because the final diagnosis was influenced by the test results [32, 33]. However, because our goal was to assess the quality of the studies from each protocol rather than the accuracy of the results, the effect of these biases may be less severe. Furthermore, even experienced clinicians are not always able to determine a specific cause in every case.
Although the gated protocol requires further refinements, our preliminary results suggest that the nongated chest pain protocol can be used as a single MDCT technique for the evaluation of acute chest pain. It allows superior pulmonary vessel and aortic contrast medium delivery along with all the advantages of contrast-enhanced CT and tailored CTA for the evaluation of the pulmonary arteries, the aorta, and the rest of the thorax.
Acknowledgments
We are grateful to Melvin Clouse, MD, for invaluable guidance, Donna Wolfe
for editorial assistance, and Clotell Forde for editorial and secretarial
assistance.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  P. M. Colletti Cardiac Imaging: Radiologists Prepare, Participate, and Publish Am. J. Roentgenol., December 1, 2007; 189(6): 1271 - 1271. [Full Text] [PDF]
|
|
|
|
|
|
P. M. Colletti Cardiac Imaging 2006 Am. J. Roentgenol., June 1, 2006; 186(6_Supplement_2): S337 - S340. [Full Text] [PDF]
|
|
| ||||||||||
