| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Nippon Telephone and Telegraph East Tohoku Hospital,
2-29-1, Yamatomachi, Wakabayashi-ku, Sendai City, Miyagi Prefecture,
Japan.
2 Department of Radiology, Tohoku University School of Medicine, 1-1, Seiryocho,
Aoba-ku, Sendai City, Miyagi Prefecture, Japan.
3 Department of Gastroenterology, Sendai City Medical Center, 5-22-1, Turugaya,
Miyagino-ku, Sendai City, Miyagi Prefecture, Japan.
Received February 12, 2002;
accepted after revision July 19, 2002.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Three types of phantoms representing the biliary system were constructed. The first phantom type was a straight tube; the second, a single tube in which the inlet and outlet diameters varied by a ratio of as much as 1:6; and the third, a tube that simulated a stricture in the biliary system and a gallstone. All experiments were repeated three times.
RESULTS. We did not observe any flow artifacts in the experiments we performed with the straight tubes. A higher rate of flow resulted in decreased signal intensity in tubes simulating bile ducts; the decreased signal was most likely to be observed on images in which the speed of flow exceeded 5 mm/sec. Flow artifacts were seen only if the ratio between the inlet and outlet diameters was 1:4 or greater. Simulations of bile duct abnormalities—such as a 50% stricture or the presence of a gallstone—did not produce any flow artifacts.
CONCLUSION. In our experiments, a flow artifact could be seen on images in which the ratio between the inlet and the outlet diameters in the phantom was equal to or greater than 1:4. This finding indicates that a flow artifact could be observed in dilated bile ducts on MRCP under clinical conditions. Knowing that a pseudo—filling defect can be caused by a flow artifact should help to prevent misinterpretation of MRCP images.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MRCP images have been reported to have potential diagnostic pitfalls that simulate or mask various abnormal conditions of the extrahepatic biliary system [10, 11]. A false-positive finding of either ductal narrowing or obstruction of the extrahepatic duct may be caused by the MRCP appearance of metallic surgical clips or intravascular coils as well as gas in the stomach or duodenum.
A flow artifact is one cause of the presence of a false filling defect at the distal bile duct on MRCP [9, 10, 12]. David et al. [12] reported that flow artifacts tend to appear in the dilated bile duct. It has previously been shown that slow flow as well as fast flow can produce signal loss in half-Fourier RARE MR imaging sequences [13]. This signal loss, to our knowledge, has not been fully investigated. We used phantoms to study the influence of the shape of the structure being imaged and the velocity flow on the appearance of flow artifacts on MRCP.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
In the first type of model, we used straight tubes with the same diameter throughout the entire length. This first model type was also used to determine the relationship between the flow velocity and the signal intensity in the phantom. Steady flow in this tube was produced using the flow apparatus that is shown schematically in Figure 2. An injector outside the bore of the MR imaging scanner supplied fluid to the model that was placed within the magnetic field. The injector regulated the flow rate. Fluid from the model was saved in the reservoir. A constant rate of flow and, therefore, a steady pressure were maintained during the experiment. A control phantom filled with a solution of cupric sulfate was placed inside the bile duct phantom.
|
Flow speed in the phantom was calculated using the injector flow rate and tube diameter. At 1.0 mm/sec intervals, flow speed was varied, ranging from approximately 1.0 to 20.0 mm/sec. This velocity range concurred with those described in experiments on bile duct function [14, 15].
The upstream inlet portion of the second model type had varied diameters of 2, 4, 8, and 12 mm to simulate bile duct sizes ranging from normal to dilated. The diameter of the narrower portion of the outlet was fixed at 2 mm, which was assumed to be the diameter of the papilla of Vater [14, 15]. The ratio between the diameter of the upstream inlet portion and the narrower outlet portion was noted, so that comparisons of the incidence of flow artifacts on MRCP images could be made.
In the third type of model, we sought to simulate various abnormalities of the bile duct system produced by stenosis or the presence of a gall-stone. This model was composed of a straight tube with a 10-mm internal diameter and a focal stenosis (inner diameter at the stenotic portion was 5 mm) and a straight tube with a 10-mm internal diameter in which a surgical specimen of gallstone (approximately a 5-mm diameter) was placed. All experiments were repeated three times.
Although the viscosity of normal bile juice varies from person to person [16] and the viscosity of the fluid affects its flow behavior, we used sterile water to represent the bile fluid for the sake of simplicity. We observed the state of flow by injecting dye solution into the phantom tube filled with water under conditions generating flow artifacts.
Imager and Scanning Parameters
All imaging was performed on a 1.5-T MR imaging system (Signa Horizon
Echospeed; General Electric Medical Systems, Milwaukee, WI). In an initial
study, we performed MRCP using a multisection half-Fourier RARE sequence with
a phased array coil. In all MRCP studies, the following parameters were
constant: TR/effective TE, infinite/90; echo-train length, 128; field of view,
200 mm; section thickness, 3 mm; number of sections, 10; section overlap, 0
mm; matrix, 256 x 192; and number of signals acquired, 1. The imaging
time was approximately 8 sec. The long axis of the phantom was placed parallel
to the long axis of the imaging table, and coronal images were obtained.
Image Analysis
In the phantom study, the relative ratio between the signal intensity in
the simulated bile duct and that in the control phantom was calculated. In the
first model type, the arbitrary area of the decreased signal was measured. In
the second type, the area upstream from the narrower outlet was measured. In
the third type, both sides of the "abnormal" portions of the
phantom were measured. The region of interest specified was greater than 2 mm
in diameter and smaller than the diameter of the phantom. The signal intensity
was measured three times, and then the three measurements were averaged.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Influence of the Ratio Between the Inlet and Outlet Diameters
Decreased signal intensity in the central part of the phantom tube or flow
artifacts were seen only in MRCP images of phantoms in which the ratio between
the inlet and outlet diameters was 1:4 or greater
(Fig. 4). Flow artifacts were
seen regardless of the rate of flow but became more pronounced as the flow
rate approached 5.0 mm/sec and were most pronounced on images in which the
flow rate exceeded 5.0 mm/sec. In MRCP images of a phantom in which the ratio
between the inlet and outlet diameters was less than 1:4, flow artifacts were
not observed, but the signal intensity in the phantom tube was observed to be
diffusely decreased compared with that in MRCP images of the control phantom.
Experiments with a dye solution showed a sharply pointed flow head near the
outlet that resembled a flow artifact (Fig.
5).
|
|
Influence of Bile Duct Disease
We did not observe any flow artifact in images of the phantom simulating
luminal stenosis. We observed the bile stone itself, but we did not observe a
flow artifact in MRCP images of the phantom with a biliary stone.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although flow-related signal loss on images obtained with a half-Fourier RARE MR imaging sequence has been previously described, we found that even a flow rate as slow as 1.0 mm/sec can decrease signal intensity [13]. This finding suggests that slow flow (i.e., as slow as the flow of bile) can produce a linear filling defect on MRCP. The cause of the decreased signal intensity is thought to be dephasing produced when the scanning plane is parallel to the magnetic field.
In our experiments, the ratio between the inlet and outlet diameters was the most important causative factor of flow artifacts. This finding suggests that in clinical practice, a flow artifact would be observed on MRCP images of a dilated bile duct (Fig. 6). David et al. [12] reported that flow artifacts tend to appear on MRCP images obtained in patients with dilated bile ducts. Results of our study are congruent with their findings.
|
We speculate that a ratio of 1:4 or greater between the inlet and outlet diameters in the phantoms caused a flow artifact for several reasons. The flow in the straight tube generally showed a plug or laminar flow. Plug flow had a linear flow head perpendicular to the wall, whereas laminar flow had a parabolic flow head (Fig. 7A,7B,7C). Therefore, even if flow in the bile duct is similar to the flow and conditions in our experiment, a filling defect would not be observed because signal loss would occur throughout the whole bile duct. However, when the ratio between the inlet and outlet diameters in the phantom equaled or exceeded 1:4, the flow head developed a sharp pointed shape. Under such conditions in the bile duct, a linear intensity decrease may occur along a course that is nearly parallel to the wall of the bile duct wall, thereby causing a flow artifact (Fig. 7A,7B,7C).
|
|
|
In previous reports, filling defects caused by flow were found in the center of the lower bile duct on axial T2-weighted MR images [9, 10, 12]. Although we did not investigate this finding, this phenomenon may also be explained as a flow artifact. The occurrence of this phenomenon may be due to a flow void caused by the scanning plane being perpendicular to the magnetic field.
On the MRCP images of the phantom simulating diseased bile ducts, we did not detect flow artifacts. Evans et al. [17, 18] showed that turbulence due to stenosis decreased the signal intensity of the flow. However, our results suggest that a flow artifact cannot be triggered by turbulence. In fact, the flow speed in our experiments did not meet the conditions required to cause turbulent flow [17, 18]. If the degree of stenosis or the size of a gallstone differs from those simulated in our phantom, flow artifacts might be detected.
Consideration should be given to whether the linear filling defect we observed could be related to factors other than flow phenomenon. Other contributory causes of linear filling defects are believed to be Mach bands, partial volume effects, and machine-dependent events. A Mach band has been described as an edge enhancement caused by the process of lateral inhibition in the retina [19]. For a filling defect to be attributed to a Mach band, the defect must be seen throughout the whole bile duct. However, we did not always observe this defect throughout the entire phantom bile duct. Therefore, we conclude that the Mach band exists independently of the linear filling defect.
A partial volume effect may be another cause of a filling defect. However, the linear filling defect always was seen at the center of the representation of the common bile duct, whereas a partial volume effect would appear at the periphery of the object being imaged. Hence, we assume that the partial volume effect exists independently of the linear filling defect.
We considered the possibility that the appearance of the linear filling defect is related to the MR imaging scanner from a specific manufacturer. However, we saw similar filling defects on images obtained on MR imaging scanners made by different manufacturers (ACS-NT, Philips Medical Systems, Shelton, CT; and Magnetom Impact/Expert, Siemens, Erlangen, Germany). Moreover, findings from other studies have indicated that a filling defect caused by flow effect can be observed on MR imaging scanners from different manufacturers (ACS-NT, Philips; and Magnetom Vision, Siemens), although the scanning sequence used in those studies was axial T2-weighted MR imaging [9, 10]. Therefore, we believe that the linear filling defect on MRCP is not related to the brand of MR imaging scanner used but that it is, instead, a flow phenomenon.
One limitation of our study is that we used a steady flow of fluid, whereas the actual flow of bile duct juice is intermittent. However, because the actual speed of bile juice flow ranges from 1.0 to 20.0 mm/sec and because we could not simulate intermittent flow, we considered that the flow artifact that we observed to be similar to an artifact that would be seen clinically.
Another technical limitation of our study concerns the shape of the phantom. We used two straight tubes. The simulation of the papilla of Vater corresponded anatomically to the connecting section of the straight tube but was not precise. In fact, the papilla of Vater displays a more complicated form; moreover, its shape may change temporarily. Therefore, we could not precisely simulate the papilla of Vater.
In addition, because bile has a more variable and often higher viscosity than our simulated bile juice, our results in a clinical situation could differ from those in our experiments. We used only water as a bile juice because every possible viscosity of bile juice is not easily reproducible with a model and because attempting to duplicate variations in viscosity was beyond the scope of our investigation.
The differentiation between a pseudo—filling defect resulting from a flow artifact and an actual filling defect resulting from disease requires careful interpretation of coronal source MR images and transverse T2-weighted MR images. The bile juice flows intermittently. Therefore, all MRCP images and other transverse MR images should not show the same pseudo—filling defect. Our results show the importance of careful interpretation of a filling defect on MRCP images in patients with dilated bile ducts. Pathologic conditions, such as a tiny bile duct stone or a tumor in the papilla of Vater, sometimes do not show true filling defects on all MRCP images and other transverse MR images. Therefore, in some patients, endoscopic retrograde cholangiopancreatography or endoscopic sonography is necessary to exclude the possibility of a true filling defect. Misdiagnosis of biliary disease can be avoided in patients whose imaging studies display pseudo—filling defects by confirming the absence of clinical symptoms.
In conclusion, our results showed that images obtained in the tubes of phantoms simulating the biliary system display flow artifacts only when the ratio between the inlet and outlet diameters was 1:4 or greater. These findings suggest that under clinical conditions, the dilated bile duct tends to cause a flow artifact. The knowledge that a pseudo—filling defect can be caused by a flow artifact should help to prevent misinterpretation of MRCP images.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  A. C. Silva, J. L. Friese, A. K. Hara, and P. T. Liu MR Cholangiopancreatography: Improved Ductal Distention with Intravenous Morphine Administration RadioGraphics, May 1, 2004; 24(3): 677 - 687. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
