(Radiology. 2000;217:115-118.)
© RSNA, 2000
Vascular and Interventional Radiology |
Suppression of Intravascular Signal on Fat-saturated Contrast-enhanced Thoracic MR Arteriograms1
Evan S. Siegelman, MD,
Riad Charafeddine, MD,
Alan H. Stolpen, MD, PhD and
Leon Axel, MD, PhD
1 From the Department of Radiology, University of Pennsylvania Medical Center, 3400 Spruce St, 1st Floor Founders—MRI, Philadelphia, PA 19104-4283. From the 1999 RSNA scientific assembly. Received June 28, 1999; revision requested August 4; revision received February 7, 2000; accepted February 22. Address correspondence to E.S.S. (e-mail: siegelm@oasis.rad.upenn.edu).
 |
ABSTRACT
|
|---|
PURPOSE: To assess the prevalence of artifactual signal intensity loss within the aortic arch and proximal branch vessels on fat-saturated contrast material–enhanced magnetic resonance (MR) arteriograms of the thoracic aorta and to hypothesize about the cause of the loss of signal intensity.
MATERIALS AND METHODS: Between January and June 1998, 105 consecutive MR arteriograms of the thoracic aorta were acquired in 103 patients at 1.5 T. Imaging included an arterial phase three-dimensional (3D) fat-saturated contrast-enhanced gradient-echo (GRE) sequence followed by a delayed two-dimensional (2D) transverse fat-saturated GRE sequence. All MR images were reviewed by two radiologists who were blinded to patient history and results of imaging studies and who evaluated the images for the presence of intraluminal loss of signal intensity in the aortic arch and the proximal branch vessels.
RESULTS: Intravascular loss of signal intensity was present in at least one vessel on 23 of the 105 arterial phase 3D studies. Seventy-one of 91 left subclavian arterial segments had loss of signal intensity on the delayed 2D studies.
CONCLUSION: Intravascular signal intensity loss can be present on contrast-enhanced fat-saturated images of the aortic arch and proximal branch vessels, particularly the left subclavian artery. This phenomenon, which is to the authors’ knowledge previously unreported and which is hypothesized to result from undesired water saturation, should not be misinterpreted as stenotic or occlusive vascular disease.
Index terms: Magnetic resonance (MR), artifact, 562.93, 912.93, 942.93 • Magnetic resonance (MR), chemical shift, 562.121414, 912.129414, 942.129414 • Magnetic resonance (MR), contrast enhancement, 562.12143, 912.12943, 942.12943 • Magnetic resonance (MR), fat suppression, 562.129415, 912.129415, 942.129415 • Magnetic resonance (MR), maximum intensity projection, 562.12142, 912.12942, 942.12942 • Magnetic resonance (MR), vascular studies, 562.12142, 912.12942, 942.12942
|
INTRODUCTION
|
|---|
Three-dimensional (3D) contrast material–enhanced magnetic resonance (MR) angiography has become the MR imaging method of choice for evaluating the thoracic aorta and its major branches (1–4). We have observed, to our knowledge a previously unreported finding of artifactual signal intensity loss within the aorta and proximal branch vessels on fat-saturated 3D MR angiograms. The purpose of this study was to assess the prevalence of this artifact among consecutive cases and to hypothesize about the cause of this signal intensity loss.
|
MATERIALS AND METHODS
|
|---|
One hundred five consecutive 3D MR angiographic examinations of the thoracic aorta were performed in 103 patients between January 1 and June 30, 1998. The patients’ ages ranged from 8 to 90 years (mean, 58 years). The majority of the studies were performed to evaluate for aortic dissection or aneurysm. All patients were imaged at 1.5 T by using "HiSpeed" gradients (Signa MRI system; GE Medical Systems, Milwaukee, Wis).
The contrast-enhanced studies were performed with a 3D dynamic fast spoiled gradient-echo (GRE) sequence (minimum repetition time msec/minimum echo time msec of <7.4/<2.5), 10°–40° flip angle, 26–44 partitions, and a 4.0–4.4-mm section thickness reconstructed at 2.0–2.2-mm increments [2x zero-fill interpolation]). The matrix used was 256–512 in the frequency-encoding direction and 128–160 in the phase-encoding direction. A bandwidth of 64 kHz was used, and the number of signals acquired was 0.5–1.0.
Contrast material (0.1–0.2 mmol of gadopentetate dimeglumine [Magnevist; Berlex Laboratories, Wayne, NJ] or gadodiamide [Omniscan; Sanofi Winthrop Pharmaceuticals, New York, NY] per kilogram of body weight) was injected rapidly by hand (approximately 1–3 mL/sec), followed by a normal saline flush. Image acquisition was initiated by the physician during contrast material injection after the patient’s circulation time was estimated. The latter was performed after briefly examining the patient and the patient’s clinical data. A timing bolus injection was not performed.
Fat saturation was performed in 99 of the 3D MR angiographic examinations by using a spectrally fat-selective inversion pulse (inversion time, 27 msec; spectral inversion at lipids, or SPECIAL, GE Medical Systems]). In the remaining six 3D MR angiographic examinations, fat saturation was not used. Images were directly acquired in the sagittal plane in 78 examinations and in the coronal plane in 27 examinations.
Following 3D MR angiography, delayed two-dimensional (2D) transverse fat-saturated spoiled GRE images were obtained through the chest. Imaging parameters included 100–250/1.4–2.1, 90° flip angle, 128–192 phase-encoding matrix, and 5–8-mm section thickness with a 1-mm intersection gap. Fat saturation was attained by using a spectrally selective saturation pulse (ChemSAT; GE Medical Systems). Both the 2D and the 3D images were acquired during suspended respiration.
Both the 3D MR angiograms and the 2D GRE images were retrospectively evaluated on an independent workstation (Advantage Windows; GE Medical Systems) by two viewers blinded to patient history and the results of imaging studies. The arterial phase volume set was reconstructed to create maximum intensity projection (MIP) images of the aortic arch and proximal branch vessels. The signal intensities of the aortic arch and normal-caliber proximal brachiocephalic artery, left common carotid artery, and left subclavian artery were evaluated by both reviewers in consensus for both the 3D MR angiographic and the 2D GRE sequences. The degree of signal intensity loss was graded on a four-point scale as follows: 0, no loss of signal intensity; 1, minimal loss of signal intensity; 2, moderate loss of signal intensity; and 3, marked loss of signal intensity (signal intensity equal to background signal intensity).
For illustrative purposes, we have included images from a patient who was examined both with and without fat-saturation techniques (Fig 1). The intravascular signal intensity loss present on the fat-saturated images in this patient was the catalyst for performing this study.
View larger version (113K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1a. Images obtained in a 47-year-old man with progressive left-upper-extremity pain and weakness. Three-dimensional MR angiography was performed to exclude a left subclavian artery stenosis or thrombosis. (a) MIP image from a fat-saturated 3D MR angiographic examination (6.0/1.2, 20° flip angle) shows moderate loss of signal intensity within the proximal segment of the left subclavian artery (arrow). (b) The patient returned 6 days later for repeat imaging. An identical 3D MR angiographic examination was performed, except that fat suppression was not used. Corresponding MIP image shows a normal left subclavian artery. (c) Delayed fat-saturated 2D GRE MR image (150/2.1, 90° flip angle) shows a signal void (straight arrow) within the left subclavian artery. Adjacent mediastinal fat (curved arrow) is not saturated. (d) Corresponding 2D GRE MR image obtained with otherwise identical imaging parameters, except that no fat-saturation pulse was used, shows normal intravascular signal intensity within the left subclavian artery. Partial signal intensity loss in the surrounding fat is secondary to opposed-phase effects.
|
|
View larger version (108K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1b. Images obtained in a 47-year-old man with progressive left-upper-extremity pain and weakness. Three-dimensional MR angiography was performed to exclude a left subclavian artery stenosis or thrombosis. (a) MIP image from a fat-saturated 3D MR angiographic examination (6.0/1.2, 20° flip angle) shows moderate loss of signal intensity within the proximal segment of the left subclavian artery (arrow). (b) The patient returned 6 days later for repeat imaging. An identical 3D MR angiographic examination was performed, except that fat suppression was not used. Corresponding MIP image shows a normal left subclavian artery. (c) Delayed fat-saturated 2D GRE MR image (150/2.1, 90° flip angle) shows a signal void (straight arrow) within the left subclavian artery. Adjacent mediastinal fat (curved arrow) is not saturated. (d) Corresponding 2D GRE MR image obtained with otherwise identical imaging parameters, except that no fat-saturation pulse was used, shows normal intravascular signal intensity within the left subclavian artery. Partial signal intensity loss in the surrounding fat is secondary to opposed-phase effects.
|
|
View larger version (138K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1c. Images obtained in a 47-year-old man with progressive left-upper-extremity pain and weakness. Three-dimensional MR angiography was performed to exclude a left subclavian artery stenosis or thrombosis. (a) MIP image from a fat-saturated 3D MR angiographic examination (6.0/1.2, 20° flip angle) shows moderate loss of signal intensity within the proximal segment of the left subclavian artery (arrow). (b) The patient returned 6 days later for repeat imaging. An identical 3D MR angiographic examination was performed, except that fat suppression was not used. Corresponding MIP image shows a normal left subclavian artery. (c) Delayed fat-saturated 2D GRE MR image (150/2.1, 90° flip angle) shows a signal void (straight arrow) within the left subclavian artery. Adjacent mediastinal fat (curved arrow) is not saturated. (d) Corresponding 2D GRE MR image obtained with otherwise identical imaging parameters, except that no fat-saturation pulse was used, shows normal intravascular signal intensity within the left subclavian artery. Partial signal intensity loss in the surrounding fat is secondary to opposed-phase effects.
|
|
View larger version (169K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1d. Images obtained in a 47-year-old man with progressive left-upper-extremity pain and weakness. Three-dimensional MR angiography was performed to exclude a left subclavian artery stenosis or thrombosis. (a) MIP image from a fat-saturated 3D MR angiographic examination (6.0/1.2, 20° flip angle) shows moderate loss of signal intensity within the proximal segment of the left subclavian artery (arrow). (b) The patient returned 6 days later for repeat imaging. An identical 3D MR angiographic examination was performed, except that fat suppression was not used. Corresponding MIP image shows a normal left subclavian artery. (c) Delayed fat-saturated 2D GRE MR image (150/2.1, 90° flip angle) shows a signal void (straight arrow) within the left subclavian artery. Adjacent mediastinal fat (curved arrow) is not saturated. (d) Corresponding 2D GRE MR image obtained with otherwise identical imaging parameters, except that no fat-saturation pulse was used, shows normal intravascular signal intensity within the left subclavian artery. Partial signal intensity loss in the surrounding fat is secondary to opposed-phase effects.
|
|
|
RESULTS
|
|---|
In the six 3D MR angiographic studies performed without fat saturation, no intravascular signal intensity loss was present within any of the 24 vascular segments. In the remaining 99 studies performed with fat saturation, 23 showed one or more vessels with some loss of signal intensity (Figs 1a, 2). The left subclavian artery was involved in 14 examinations (mode = 1), the aortic arch in 10 (mode = 1), the left common carotid artery in two (mode = 1), and the brachiocephalic artery in one (mode = 1).
View larger version (94K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2. MIP image from a fat-saturated 3D MR angiographic examination (6.0/1.3, 30° flip angle) in a 42-year-old woman shows mild loss of signal intensity within the proximal segments of the left subclavian (straight arrow) and left common carotid (curved arrow) arteries. In this case, these regions of signal intensity loss are subtle and likely explain why this phenomenon has not previously been reported. A third segment of artifactual signal intensity loss is present within the distal descending thoracic aorta (arrowhead).
|
|
Fourteen of the 105 studies did not have 2D GRE images that were interpretable because 2D GRE imaging was not performed, the superior aspect of the chest was not imaged, or the images were degraded by marked respiratory artifact. Of the remaining 91 studies, 72 (79%) showed signal intensity loss in at least one vessel (Figs 1c, 3b). Some degree of signal intensity loss was present in the left subclavian artery in 71 studies (mode = 2), in the left common carotid artery in 28 (mode = 1), in the brachiocephalic artery in 11 (mode = 1), and in the aortic arch in nine (mode = 1).
View larger version (124K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3a. Artifactual signal intensity loss secondary to two different mechanisms in the subclavian arteries in a 49-year-old man. (a) MIP image from a 3D MR angiographic examination (6.0/1.2, 30° flip angle) shows no intravascular signal intensity (arrow) within the expected course of the middle and peripheral parts of the right subclavian artery secondary to T2-shortening effects of concentrated gadolinium-containing contrast material in the adjacent subclavian vein. This artifact typically appears in the subclavian arterial segment located at the midclavicular level and resolves on delayed images when concentrated gadolinium-containing contrast material is no longer present in the subclavian vein (not shown). Aortic arch and proximal branches are normal. (b) Transverse 2D GRE MR image (150/1.5, 90° flip angle) through the proximal branches of the aorta shows marked signal intensity loss (straight arrow) within the left subclavian artery. Note the lack of suppression of the surrounding mediastinal fat (curved arrow).
|
|
View larger version (127K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3b. Artifactual signal intensity loss secondary to two different mechanisms in the subclavian arteries in a 49-year-old man. (a) MIP image from a 3D MR angiographic examination (6.0/1.2, 30° flip angle) shows no intravascular signal intensity (arrow) within the expected course of the middle and peripheral parts of the right subclavian artery secondary to T2-shortening effects of concentrated gadolinium-containing contrast material in the adjacent subclavian vein. This artifact typically appears in the subclavian arterial segment located at the midclavicular level and resolves on delayed images when concentrated gadolinium-containing contrast material is no longer present in the subclavian vein (not shown). Aortic arch and proximal branches are normal. (b) Transverse 2D GRE MR image (150/1.5, 90° flip angle) through the proximal branches of the aorta shows marked signal intensity loss (straight arrow) within the left subclavian artery. Note the lack of suppression of the surrounding mediastinal fat (curved arrow).
|
|
In the representative patient after the study period, signal intensity loss was present in the left subclavian artery with 3D MR angiographic and 2D GRE sequences with fat saturation but not when the sequences were repeated without fat saturation (Fig 1).
|
DISCUSSION
|
|---|
To our knowledge, this phenomenon of water suppression at fat-saturated imaging has only recently been reported for spinal MR imaging (5) but not for contrast-enhanced MR angiography. Intravascular signal intensity loss on contrast-enhanced MR angiograms (obtained either with or without fat saturation) can occur secondary to adjacent stents, clips, or concentrated gadolinium-containing contrast material (Fig 3a) (6–8). The failure of fat to be saturated and, thus, mimic disease with fat-saturated fast spin-echo sequences and/or fat-saturated contrast-enhanced T1-weighted imaging can occur in locations that are adjacent to the lung bases (9), gas-containing bowel segments (10), or paranasal sinuses (11,12).
Only one of the cited studies mentioned the presence of water suppression (5) perhaps because regions of low signal intensity can be readily overlooked on fat-saturated T2-weighted images and with nonvascular applications of contrast-enhanced T1-weighted imaging. However, loss of intravascular signal intensity on 3D MR angiograms can be readily detected in many instances (Figs 1a, 2). Two imaging features that suggest that the loss of signal intensity is artifactual are the lack of luminal narrowing (the vessels we analyzed were of normal caliber) and the absence of collateral vessels. Truly stenotic or occluded proximal branch vessels also reveal decreased or absent signal intensity on contrast-enhanced MR angiograms obtained without fat saturation. Because of this artifact, we no longer routinely use fat saturation when performing contrast-enhanced MR angiographic studies of the chest.
Although we did not specifically evaluate the adjacent mediastinal fat, its lack of suppression was apparent on many of the 3D MR angiograms and 2D GRE images (Figs 1c, 3b). We hypothesize that magnetic susceptibility effects from aerated lung result in frequency shifts in adjacent fat- and water-containing tissues. Thus, when a spectrally selective fat-saturation (either conventional or inversion-recovery) technique is used, regions of water, and not adjacent fat, may be saturated. When a non–spectrally selective fat-saturation sequence (such as short inversion time inversion-recovery, or STIR, imaging) is used, the phenomenon of water suppression is not reported (5). This hypothesis has been confirmed with phantom studies (13).
The left subclavian artery exhibits the greatest likelihood of signal intensity loss. This may be related to its proximity to the medial aspect of the left upper lobe of the lung. The likely reason we observed a greater degree of signal intensity loss on the 2D GRE images compared with that on the 3D MR angiograms is because the degree of water suppression is less complete with the inversion-recovery technique used for the 3D images (because the inversion time of 27 msec is not optimal for water suppression) when compared with the conventional chemical-shift saturation pulse used for the 2D images.
In conclusion, artifactual signal intensity loss in intrathoracic vessels can occur with fat-saturated contrast-enhanced 3D MR angiographic and 2D GRE sequences and may result in false-positive diagnoses of stenotic or occlusive vascular disease.
|
FOOTNOTES
|
|---|
Abbreviations: GRE = gradient echo,
MIP = maximum intensity projection,
3D = three-dimensional,
2D = two-dimensional
Author contributions: Guarantor of integrity of entire study, E.S.S.; study concepts, E.S.S.; study design, E.S.S., R.C.; definition of intellectual content, all authors; literature research, E.S.S., R.C.; clinical studies, E.S.S., A.H.S., L.A.; data acquisition and analysis, E.S.S., R.C.; manuscript preparation, E.S.S.; manuscript editing, E.S.S., A.H.S., L.A.; manuscript review, all authors.
|
REFERENCES
|
|---|
-
Krinsky GA, Rofsky NM, DeCorato DR, et al. Thoracic aorta: comparison of gadolinium-enhanced three-dimensional MR angiography with conventional MR imaging. Radiology 1997; 202:183-193.[Abstract/Free Full Text]
-
Ho VB, Prince MR. Thoracic MR aortography: imaging techniques and strategies. RadioGraphics 1998; 18:287-309.[Abstract]
-
Prince MR, Narasimham DL, Jacoby WT, et al. Three-dimensional gadolinium-enhanced MR angiography of the thoracic aorta. AJR Am J Roentgenol 1996; 166:1387-1397.[Abstract/Free Full Text]
-
Leung DA, Debatin JF. Three-dimensional contrast-enhanced magnetic resonance angiography of the thoracic vasculature. Eur Radiol 1997; 7:981-989.[Medline]
-
Nakatsu M, Hatabu H, Itoh H, et al. Comparison of short inversion time recovery (STIR) and fat-saturated (Chemsat) techniques for background fat intensity suppression in cervical and thoracic MR imaging. J Magn Reson Imaging 2000; 11:56-60.[Medline]
-
Krinsky G, Rofsky NM. MR angiography of the aortic arch vessels and upper extremities. Magn Reson Imaging Clin N Am 1998; 6:269-292.[Medline]
-
Amano Y, Gemma K, Kawamata H, Kumazaki T. Intraluminal signal intensity of iliac artery stents investigated by contrast-enhanced three-dimensional MR angiography. Comput Med Imaging Graph 1998; 22:9-12.[Medline]
-
Prince MR. Gadolinium-enhanced MR aortography. Radiology 1994; 191:155- 164.[Abstract/Free Full Text]
-
Yoshimitsu K, Varma DG, Jackson EF. Unsuppressed fat in the right anterior diaphragmatic region on fat-suppressed T2-weighted fast spin-echo MR images. J Magn Reson Imaging 1995; 5:145-149.[Medline]
-
Siegelman ES, Outwater EK. MR imaging techniques of the liver. Radiol Clin North Am 1998; 36:263-286.[Medline]
-
Anzai Y, Lufkin RB, Jabour BA, Hanafee WN. Fat-suppression failure artifacts simulating pathology on frequency-selective fat-suppression MR images of the head and neck. AJNR Am J Neuroradiol 1992; 13:879-884.[Abstract]
-
Borges AR, Lufkin RB, Huang AY, Farahani K, Arnold AC. Frequency-selective fat suppression MR imaging: localized asymmetric failure of fat suppression mimicking orbital disease. J Neuroophthalmol 1997; 17:12-17.[Medline]
-
Axel L, Kolman L, Charafeddine R, Hwang S, Stolpen AH. Origin of a signal loss artifact in fat suppression MRI. Radiology; (in press).
This article has been cited by other articles:
|
|
|
|
 
L. Axel, L. Kolman, R. Charafeddine, S. N. Hwang, and A. H. Stolpen
Origin of a Signal Intensity Loss Artifact in Fat-Saturation MR Imaging
Radiology,
December 1, 2000;
217(3):
911 - 915.
[Abstract]
[Full Text]
|
|
|
TOP
ABSTRACT
INTRODUCTION
