(Radiology. 1999;210:113-123.)
© RSNA, 1999
Gas at Abdominal US: Appearance, Relevance, and Analysis of Artifacts
Stephanie R. Wilson, MD1,3,
Peter N. Burns, PhD2,4,
Lana M. Wilkinson, BSc2,4,
David Hope Simpson, MSc2,4 and
Derek Muradali, MD1,3
1 Departments of Medical Imaging (S.R.W., D.M.)
2 Medical Biophysics (P.N.B., L.M.W., D.H.S.)
3 University of Toronto; the Department of Medical Imaging, The Toronto Hospital (S.R.W., D.M.), 200 Elizabeth St, Toronto M5G 2C4, Ontario, Canada
4 Department of Imaging Research, Sunnybrook Health Science Centre (P.N.B., L.M.W., D.H.S.), Toronto.
 |
Abstract
|
|---|
PURPOSE: To describe the spectrum of ultrasonographic (US) appearances of intraluminal gas, including two clinically relevant gas artifacts.
MATERIALS AND METHODS: Observations were made in patients and reproduced in an animal model, an ex vivo gut preparation, and a tissue-mimicking phantom. Appearances were classified according to a physical model of the interaction between sound and collections of gas.
RESULTS: Free bubbles of gas appeared as bright echogenic foci extending artifactually owing to lateral and axial blooming. This causes bubbles that abut the gut wall to enhance the layer one echo, which corresponds to the interface between the mucosa and the luminal contents. Such bubbles can also falsely appear to be within the gut wall itself owing to elevation averaging and thereby cause the artifact pseudo–pneumatosis intestinalis. Isolated groups of small bubbles created a characteristic periodicity and tapering of the distal echo pattern. In the extreme case, in which many such echoes are superimposed, "dirty shadowing" occurs. A contiguous pocket of gas may cause the gut wall to appear artifactually thickened (ie, pseudo–thickened gut). This was shown to be a form of mirror image artifact.
CONCLUSION: Classification of the effects of gas on US images according to their physical characteristics may aid in their interpretation. Appreciating two previously undescribed artifacts, pseudo–pneumatosis intestinalis and pseudo–thickened gut, will improve the usefulness of abdominal US.
Index terms: Gastrointestinal tract, US, 74.12989, 75.12989 • Intestines, cysts, 74.782, 75.782 • Ultrasound (US), artifact, 74.1298, 75.1298 • Ultrasound (US), experimental, 74.12989, 75.12989
|
Introduction
|
|---|
Artifacts associated with gas in the lumen of the gut, which are familiar to all sonographers, occur whenever the ultrasound beam encounters gas in the scanning field. In fact, "dirty shadowing" and reverberation artifacts distal to a bright echogenic focus of gas are encountered so frequently that they are considered to be a normal observation at abdominal and pelvic examinations. Although gas is a frequent cause of substantial degradation of the ultrasonographic (US) image and may obscure underlying structures, it does not necessarily have a negative influence on the US examination. The identification of gas in an abnormal location such as the gallbladder or the kidney may correctly lead to a diagnosis of emphysematous cholecystitis or emphysematous nephritis. Similarly, gas within a fluid collection suggests an abscess with gas-forming organisms, and gas in the peritoneal cavity or retroperitoneum has a high association with gut perforation.
Gas bubbles may also be located abnormally within the wall of the gut in a condition called pneumatosis intestinalis, which is commonly associated with necrotizing enterocolitis in the pediatric population or with gut necrosis in adults. This observation is of great importance as a negative prognostic indicator. Identification of pneumatosis intestinalis on US images (Fig 1) requires meticulous technique to appreciate the circumferential bright echogenic foci within the gut wall that represent the gas bubbles. However, pneumatosis intestinalis may be mimicked by pockets of intraluminal gas in the nondependent part of the gut lumen that become temporarily trapped within mucosal folds (Fig 2). We have observed that bubbles of intraluminal gas may appear falsely to be lying within the gut wall itself; this produces an artifact that we refer to as pseudo–pneumatosis intestinalis (Fig 3). Gas in the lumen of the gut also can result in an US image that falsely depicts gut wall thickening; we refer to this artifact as pseudo–thickened gut (Fig 4). We believe that the artifacts pseudo–pneumatosis intestinalis and pseudo–thickened gut are unfamiliar to most sonographers and have the potential to cause erroneous interpretations of US images and thus lead to negative patient outcome.
View larger version (172K):
[in this window]
[in a new window]
|
Figure 1. True pneumatosis intestinalis in a 50-year-old woman with ischemic gut associated with a closed loop obstruction. Long-axis US image of a fluid-filled loop of the small bowel (L) shows circumferential bright echogenic foci (arrows) arising from gas bubbles within the gut wall.
|
|
View larger version (170K):
[in this window]
[in a new window]
|
Figure 2. Gas within the mucosal folds mimicking pneumatosis intestinalis in a 36-year-old woman. Cross-sectional US image of the gastric antrum shows multiple bright echogenic foci (arrows), which suggest intramural gas predominantly in the nondependent wall of the stomach (L). A second US image obtained a few seconds later showed a normal multilayered appearance, which confirmed the artifactual nature of the appearance.
|
|
View larger version (167K):
[in this window]
[in a new window]
|
Figure 3. Pseudo–pneumatosis intestinalis in an 86-year-old man with abdominal pain and distention due to a gallstone ileus. Long-axis, 3.5-MHz US image of a dilated fluid-filled loop (L) of the small bowel shows bright echogenic foci related to the gut wall; these foci raised the possibility of pneumatosis intestinalis. An obstructing gallstone was seen on another view. Plain radiographs and computed tomographic images did not confirm pneumatosis.
|
|
View larger version (174K):
[in this window]
[in a new window]
|
Figure 4. Pseudo–thickened gut in a 50-year-old man in the intensive care unit with nonspecific abdominal pain. Sagittal US image of the lower part of the abdomen shows a multilayered appearance, which suggests a loop of thickened gut on the long axis. This is an artifact due to gas at the gut wall interface (arrows), which was confirmed on an identical orthogonal view.
|
|
The familiar artifacts of dirty shadowing and small reverberation artifacts distal to an echogenic focus of gas represent a small portion of the potential US appearances of gas artifacts. The purposes of this study were to (a) demonstrate the complete spectrum of appearances of intraluminal gas at US and reproduce this spectrum in an ex vivo model and in tissue-mimicking material, (b) propose physical explanations for these appearances, including the previously undescribed pseudo–pneumatosis intestinalis and pseudo–thickened gut, and (c) suggest guidelines for the recognition and differentiation of the artifacts in clinical practice.
|
MATERIALS AND METHODS
|
|---|
The US appearances of gas in vivo in humans and pigs, in an ex vivo porcine stomach, and in a tissue-mimicking phantom were documented.
In Vivo Studies
Baseline observations of the appearance of gas in the gut were made during 350 consecutive abdominal or pelvic US examinations performed in adult patients who were referred for evaluation of gastrointestinal pathologic conditions during 1 year. Patients suspected of having appendicitis and diverticulitis and those with complications of inflammatory bowel disease were regularly included in this patient complement. Our routine protocol includes a peritoneal cavity survey, with a more detailed assessment of the region of interest performed with a higher frequency linear or curvilinear US transducer.
US imaging was performed in 15 anesthetized pigs (average weight, 20 kg) from an unrelated study by using a US system similar to that used in the other study. The animal protocols used in this study were approved by the institutional review board and conformed to American Veterinary Medical Association guidelines. The sonograms in the pigs were obtained by using C4-2 and C7-4 curvilinear US transducers (Advanced Technology Laboratories, Bothell, Wash); detailed assessment was performed by using an L10-5 linear-array transducer (HDI 3000; Advanced Technology Laboratories). Observations were made on all visible gut so as to document the different appearances that we could attribute to gas pockets and gas bubbles within the pigs' intestinal tracts.
Ex Vivo Model
The porcine stomach, which is similar in gross anatomy and histologic features to the human stomach and has a good reservoir capacity, was selected as the ex vivo model for investigating the US appearances of intraluminal gas. The fresh stomach of one of the pigs from the in vivo studies was resected at the gastroesophageal and gastroduodenal junctions. Six-millimeter-diameter tubing was attached in an airtight manner to both the proximal and distal resection margins of the stomach. The proximal margin of the stomach was placed in a dependent position and slowly filled with degassed water. The stomach was distended to its maximal diameter and submerged in a tank of degassed water. The anterior gastric wall was imaged about 2 cm from the face of a fixed, 7.5-MHz linear-array transducer (UM9 HDI; Advanced Technology Laboratories). A 5-F catheter was inserted through the distal port into the lumen of the stomach. The transducer was adjusted so as to image a vertical plane that included the catheter tip and the gastric wall (Fig 5). Gas was injected through the catheter so as to produce a single bubble, a stream of bubbles, and a bolus of up to 10 mL of gas.
View larger version (52K):
[in this window]
[in a new window]
|
Figure 5. Magnified baseline pig sonogram of the gastric wall with a rugal fold (F) shows five distinct layers of alternating echogenicity. The innermost echo (arrows), which corresponds to the interface between the fluid and the mucosal surface, is white. The outermost echo is indicated by the arrowheads. Progressing from the lumen to the serosa, the wall layers are white, black, white, gray, and white. The catheter is visible as the long white line on the right side of the image. The marker interval is 5 mm.
|
|
Images of the bubbles were obtained at the catheter tip and as the bubbles ascended toward the gastric wall. The various appearances produced by the gas on contact with the gastric wall were recorded. To investigate the acoustical effects of larger volumes of gas, a 10-mL bolus of gas was injected and allowed to sit in the stomach for several minutes before imaging. Imaging sequences were reviewed as cine loops and recorded onto a magnetic-optical disk. At completion of imaging, the gastric wall was resected and examined histopathologically, and the actual wall thickness was measured.
Finally, contiguous pockets of gas more than 1 cm in diameter were produced beneath the wall in the excised stomach and in the tissue-equivalent material, and the wall was scanned from above and photographed from below for correlation with the US findings.
Tissue-mimicking Phantom
A block of tissue-mimicking material composed of an agar gel matrix, in which were suspended Biogel beads less than 40 µm in diameter (Bio-Rad Laboratories, Richmond, Calif), polyacrylamide, and graphite powder to induce attenuation, was constructed. A horizontal cylindrical void was created in the block and connected to a flow system of degassed water, into which single bubbles were injected by using a 22-gauge needle.
The bubbles were scanned through the gel by using an L10-5 linear-array transducer (UM9 HDI unit; Advanced Technology Laboratories) and confirmed to be in contact with the vessel wall. At a constant depth of 4.4 cm, transverse images of the vessel and intraluminal bubbles were acquired at 50°–140° angles at 10° intervals to the vessel void. Images were recorded digitally onto a magnetic-optical disk. The apparent distance between the center of the bubble and the vessel wall was measured on the offline image by using calibrated analysis software (NIH Image; National Institutes of Health, Bethesda, Md).
|
RESULTS
|
|---|
Baseline findings of the appearances associated with gas on abdominal sonograms showed the gas itself as an area of bright echogenicity that varied in size from a small, discrete focus to a continuous, highly echogenic band. Deep to the bright echogenicity, we observed a variety of appearances in two basic forms. The first form was that of large gas pockets that showed a chaotic display of continuous echoes extending throughout the visible field. We referred to this as dirty shadowing. The straight margins of the dirty shadow corresponded to the edges of the gas pocket. The shadow lacked interpretable information and seemed to originate from a location other than the structures of the area of insonation. The other appearance, that of more discrete foci of echogenicity, had a distal linear configuration of variable length, with periodic appearances of horizontal, parallel lines. These lines varied in both size and shape; they could be short or long and tapered or extended.
At baseline US of the gastric wall in the ex vivo porcine stomach, five distinct layers of alternating echogenicity that resulted from layered acoustic interfaces within the structure (Fig 5) were revealed. The relationship of these layers to the histomorphologic features of the stomach wall remains controversial (1,2). We labeled these acoustic layers in accordance with the system of Kimmey et al (1) as follows: layer one, the innermost hyperechoic layer, represents the interface between the intraluminal fluid and the mucosa; layer two is hypoechoic and represents the deep mucosa; layer three is hyperechoic and represents the submucosa plus the acoustic interface between the submucosa and the muscularis propria; layer four is hypoechoic and represents the muscularis propria minus the acoustic interface between the submucosa and the muscularis propria; and layer five is hyperechoic and represents the serosa and the subserosal fat.
On the basis of our observations in vivo, in ex vivo porcine stomach, and in the phantom, gas in the US scanning field produced image artifacts in three broad categories: (a) those related to a single bubble, (b) those related to a collection of multiple bubbles, and (c) those related to a contiguous pocket of gas without bubble structure. These artifacts varied according to the location, volume, and form of the bubbles. Artifacts were observed within the gastric wall, distal to a normal-appearing gastric wall, and distal to a gastric wall with an appearance that had changed artifactually. These observations are tabulated according to origin and US appearance in the Table.
Elliptical Bubble Echo with Bright Echogenic Focus
US appearance.—In the ex vivo model, the injected bubbles appeared as hyperechoic foci exiting from the catheter tip and ascending through the surrounding water. The appearance of the individual bubbles varied according to their size. Tiny bubbles were seen as hyperechoic spherical foci (Fig 6a). As they made contact with the gut wall, they appeared to roll out of the scanning plane. In contrast, larger bubbles were stronger reflectors, with echoes that extended laterally and axially and assumed an elliptical shape (Fig 6b).
View larger version (97K):
[in this window]
[in a new window]
|
Figure 6a. US appearances of free bubbles. (a) Sonogram shows tiny bubbles (small arrow) that appear as small echogenic foci. Larger bubbles (large arrow) appear brighter than the smaller bubbles and extend in both the axial and lateral directions. (b) However, the larger bubbles (arrow) extend across the effective width of the ultrasound beam; this is especially visible in the far field of the transducer. (c) Lateral blooming. The ultrasound beam does not have a perfectly sharp boundary, but rather it has tapered sensitivity in a lateral direction. Stronger targets such as bubbles are therefore artifactually seen as extending in a lateral direction at imaging (Fig 6 continues).
|
|
View larger version (72K):
[in this window]
[in a new window]
|
Figure 6b. US appearances of free bubbles. (a) Sonogram shows tiny bubbles (small arrow) that appear as small echogenic foci. Larger bubbles (large arrow) appear brighter than the smaller bubbles and extend in both the axial and lateral directions. (b) However, the larger bubbles (arrow) extend across the effective width of the ultrasound beam; this is especially visible in the far field of the transducer. (c) Lateral blooming. The ultrasound beam does not have a perfectly sharp boundary, but rather it has tapered sensitivity in a lateral direction. Stronger targets such as bubbles are therefore artifactually seen as extending in a lateral direction at imaging (Fig 6 continues).
|
|
View larger version (13K):
[in this window]
[in a new window]
|
Figure 6c. US appearances of free bubbles. (a) Sonogram shows tiny bubbles (small arrow) that appear as small echogenic foci. Larger bubbles (large arrow) appear brighter than the smaller bubbles and extend in both the axial and lateral directions. (b) However, the larger bubbles (arrow) extend across the effective width of the ultrasound beam; this is especially visible in the far field of the transducer. (c) Lateral blooming. The ultrasound beam does not have a perfectly sharp boundary, but rather it has tapered sensitivity in a lateral direction. Stronger targets such as bubbles are therefore artifactually seen as extending in a lateral direction at imaging (Fig 6 continues).
|
|
View larger version (13K):
[in this window]
[in a new window]
|
Figure 6d. Figure 6 (continued). (d) Axial blooming. Similar to the production lateral blooming, the ultrasound pulse does not have a uniform amplitude over its duration. Axial blooming is generally less severe than is lateral blooming.
|
|
Physical analysis.—A gas bubble produces an echo that is approximately 1014 times stronger than the echo produced by typical soft tissue of equivalent volume (3). Because the ultrasound beam does not have a perfectly sharp boundary like, for example, a laser beam of light, but rather the intensity of its lateral margin is tapered more like that of a flashlight beam, the effect of increasing the echo strength of a target is that the echo is detected over a greater lateral distance than is a weaker echo from the same target. This is known as lateral blooming (Fig 6c). In an analogous way, the axial extent of the resolution cell on the US image is defined as a tapering sensitivity along the direction of the ultrasound beam so that proximal and distal extension of the echogenic target (ie, the bubble) also occurs. This is known as axial blooming (Fig 6d), which in most circumstances is less severe than lateral blooming because of the more sharply defined axial sensitivity pattern of the transmitted pulse. Thus, the bubble echo is brighter because of the mismatch of the acoustic impedance and larger because of lateral and axial blooming. The extent of the blooming artifact increases with bubble size and with increasing lateral width of the ultrasound beam.
Layer One Enhancement
US appearance.—At contact with the gut wall, some bubbles caused enhancement of the echo in gastric wall layer one (Fig 7a). Enhancement of the gastric wall layer one echo increased with increasing bubble diameter. This enhancement diffused beyond the margin of the bubble in the direction of the lateral plane and caused slight augmentation of layer one in the axial direction. The same effect was seen in the tissue-equivalent phantom when a single bubble was trapped in contact with the upper wall. The junction between the water and the surface of the tissue-mimicking material produced a thin, linear echogenic interface that was similar to gastric wall layer one. This interface became enhanced when the bubble was in contact with the wall (Fig 7b). Augmentation of the echo in the lateral and axial directions was also seen; the lateral spreading was more pronounced.
View larger version (126K):
[in this window]
[in a new window]
|
Figure 7a. Layer one enhancement at US. (a) On the sonogram, at the point where the bubble is in contact with the gastric wall (between the two arrows), there is increased echogenecity of layer one (arrowhead). (b) The presence of a bubble (B) abutting the wall of the flow channel in the phantom causes enhancement of the layer one echo (arrows). (c) Physical explanation. The tissue-fluid interface is replaced by the more echogenic tissue-gas interface.
|
|
View larger version (160K):
[in this window]
[in a new window]
|
Figure 7b. Layer one enhancement at US. (a) On the sonogram, at the point where the bubble is in contact with the gastric wall (between the two arrows), there is increased echogenecity of layer one (arrowhead). (b) The presence of a bubble (B) abutting the wall of the flow channel in the phantom causes enhancement of the layer one echo (arrows). (c) Physical explanation. The tissue-fluid interface is replaced by the more echogenic tissue-gas interface.
|
|
View larger version (21K):
[in this window]
[in a new window]
|
Figure 7c. Layer one enhancement at US. (a) On the sonogram, at the point where the bubble is in contact with the gastric wall (between the two arrows), there is increased echogenecity of layer one (arrowhead). (b) The presence of a bubble (B) abutting the wall of the flow channel in the phantom causes enhancement of the layer one echo (arrows). (c) Physical explanation. The tissue-fluid interface is replaced by the more echogenic tissue-gas interface.
|
|
Physical analysis.—Layer one echo enhancement (Fig 7) occurs as the tissue-fluid interface of the gastric wall is replaced by the tissue-gas interface of the bubble in contact with the wall surface (Fig 7c). Because the acoustic impedance mismatch between the soft tissue of the gastric wall and gas is much greater than that between the gastric wall and water, the echo produced at the tissue-gas interface is stronger. Lateral and axial blooming account for the extension of the echo in these two directions. The fact that the phantom has no wall between the fluid and tissue supports the hypothesis of Kimmey et al (1) that gastric wall layer one does not represent a true anatomic structure, but rather it represents the discontinuity of acoustic impedance at the interface between the intraluminal fluid and the gastric wall. Thus, the effect of a bubble on the gut wall is the creation of an echo that is both brighter because of the acoustic impedance mismatch and larger because of lateral and axial blooming.
Pseudo–Pneumatosis Intestinalis
US appearance.—In the in vivo model, we observed that in some cases, in addition to causing layer one enhancement, the echoes from bubbles actually appeared to lie within the gastric wall itself. While some bubble echoes extended just beyond layer one, others appeared to lie completely within layers two or three. We refer to this false appearance of gas in the gut wall as pseudo–pneumatosis intestinalis (Fig 8a).
View larger version (77K):
[in this window]
[in a new window]
|
Figure 8a. Pseudo–pneumatosis intestinalis artifact at US. (a) Sonogram shows enhancement of layer one (curved arrows) and a central bubble (straight arrows) that appears artifactually to lie within the superficial layer of the gut wall (magnified view inset). (b) The vessel phantom, which does not have a wall, was scanned obliquely at a series of angles (90°, 110°, 120°, and 130°) to simulate the pseudo–pneumatosis intestinalis artifact. The bubble (arrows) is falsely depicted as lying further and further from its true location in the phantom lumen as the angulation increases. (c) US images of a bubble in contact with the mucosal surface of the stomach scanned at comparable angles show similar artifactual migration of the bubble echo (arrows) into the wall (Fig 8 continues).
|
|
View larger version (107K):
[in this window]
[in a new window]
|
Figure 8b. Pseudo–pneumatosis intestinalis artifact at US. (a) Sonogram shows enhancement of layer one (curved arrows) and a central bubble (straight arrows) that appears artifactually to lie within the superficial layer of the gut wall (magnified view inset). (b) The vessel phantom, which does not have a wall, was scanned obliquely at a series of angles (90°, 110°, 120°, and 130°) to simulate the pseudo–pneumatosis intestinalis artifact. The bubble (arrows) is falsely depicted as lying further and further from its true location in the phantom lumen as the angulation increases. (c) US images of a bubble in contact with the mucosal surface of the stomach scanned at comparable angles show similar artifactual migration of the bubble echo (arrows) into the wall (Fig 8 continues).
|
|
View larger version (90K):
[in this window]
[in a new window]
|
Figure 8c. Pseudo–pneumatosis intestinalis artifact at US. (a) Sonogram shows enhancement of layer one (curved arrows) and a central bubble (straight arrows) that appears artifactually to lie within the superficial layer of the gut wall (magnified view inset). (b) The vessel phantom, which does not have a wall, was scanned obliquely at a series of angles (90°, 110°, 120°, and 130°) to simulate the pseudo–pneumatosis intestinalis artifact. The bubble (arrows) is falsely depicted as lying further and further from its true location in the phantom lumen as the angulation increases. (c) US images of a bubble in contact with the mucosal surface of the stomach scanned at comparable angles show similar artifactual migration of the bubble echo (arrows) into the wall (Fig 8 continues).
|
|
View larger version (13K):
[in this window]
[in a new window]
|
Figure 8d. Figure 8 (continued). (d) The principle of elevation averaging, which is the cause of the pseudo–pneumatosis intestinalis artifact. The bubble, an echogenic target, lies outside the scanning plane but within the elevation thickness of the ultrasound beam. The appearance of the bubble will be misregistered on the image within the scanning plane. (e) Graph shows the measured displacement of the bubble due to elevation averaging compared with the displacement of the bubble predicted by using geometric theory. The theoretic line suggests an effective elevation width of the beam of 2.2 mm. The true elevation width at the focus of this linear-array transducer is approximately 2.0–3.0 mm.
|
|
View larger version (22K):
[in this window]
[in a new window]
|
Figure 8e. Figure 8 (continued). (d) The principle of elevation averaging, which is the cause of the pseudo–pneumatosis intestinalis artifact. The bubble, an echogenic target, lies outside the scanning plane but within the elevation thickness of the ultrasound beam. The appearance of the bubble will be misregistered on the image within the scanning plane. (e) Graph shows the measured displacement of the bubble due to elevation averaging compared with the displacement of the bubble predicted by using geometric theory. The theoretic line suggests an effective elevation width of the beam of 2.2 mm. The true elevation width at the focus of this linear-array transducer is approximately 2.0–3.0 mm.
|
|
A similar effect could be produced in the tissue-equivalent phantom by rotating the transducer 90° to scan across the vessel section and tilting it so that the angle between the scanning plane and the axis of the vessel was progressively decreased. An individual bubble appeared to migrate into the tissue material, even though the longitudinal image confirmed that the bubble really lay within the vessel lumen. Larger angles of deviation from the normal plane were associated with larger distances between the image of the bubble and the vessel void where the bubble was actually located (Fig 8b). In correlating these observations with the in vitro image findings, a similar angulation of the transducer enabled clear visualization of a single bubble migrating into the stomach wall as the transducer angle was changed (Fig 8c).
Physical analysis.—The results of the tissue-mimicking phantom experiment suggest that the pseudo–pneumatosis intestinalis artifact is the consequence of a geometric effect that is dependent on the size of the angle between the scanning plane and the bubble-wall interface. This can be understood by considering the finite thickness of the ultrasound beam in the direction orthogonal to the scanning plane known as the elevation, or section-thickness, direction. The beam of most US arrays is considerably broader in the elevation direction than in the lateral and axial directions, but this is usually not apparent to the sonographer, except as a loss of contrast at imaging of, for example, a cyst with a diameter that is less than that of the effective section thickness. This is caused by the superimposition of echoes from the outside of the cyst (but still within the imaging section) to its echo-poor interior. This artifact is analogous to the volume-averaging effects that are frequently observed on computed tomographic and magnetic resonance images.
In the unusual circumstance in which an echogenic target lies outside the scanning plane but within the elevation thickness of the ultrasound beam, the target will be misregistered on the image within the scanning plane (Fig 8d). Again, because the beam sensitivity tapers in the elevation direction rather than ends abruptly, the effective section thickness is greater for a bright target such as a bubble. Thus, elevation averaging is the cause of the pseudo–pneumatosis intestinalis artifact. For a given angle and extent of the beam in the elevation direction, a simple geometric model can help to predict the displacement of the bubble that will be apparent on the image as a function of angle. Figure 8e shows a comparison between the displacement of the bubble predicted on the basis of the simple geometric theory and the displacement measured in the tissue-equivalent phantom, which supports the proposed explanation.
Tapered Periodic Echo and Comet Tail Artifact
US appearance.—A small collection of bubbles in the excised stomach (Fig 9a) produced a column of multiple, equally spaced echoes parallel to the face of the transducer. The shape of this column varied according to the size and distribution of the individual bubbles. A single layer of tiny bubbles produced a V-shaped column, which created the classic appearance of a comet tail owing to tapered reverberation (Fig 9b).
View larger version (91K):
[in this window]
[in a new window]
|
Figure 9a. Bubble clusters. (a) Photograph of a small bubble cluster (arrows). (b) Sonogram shows a small bubble cluster that gives rise to a tapered distal artifact (arrows) with periodic echogenicity. Longer echoes occur when the bubble is aligned exactly along the axis of the ultrasound beam. (c) Sonogram shows multiple bubble clusters (arrows) that have wider, less periodic multiple echoes with lengths that depend on the size and alignment of the bubbles with respect to the ultrasound beam axis. (d) Physical explanation. A simple triad of bubbles (T) produces the periodic echoes (E) of the comet-tail artifact. Multiple clusters produce an overlying series of similar echo patterns, which eventually merge into a continuous tail (C).
|
|
View larger version (75K):
[in this window]
[in a new window]
|
Figure 9b. Bubble clusters. (a) Photograph of a small bubble cluster (arrows). (b) Sonogram shows a small bubble cluster that gives rise to a tapered distal artifact (arrows) with periodic echogenicity. Longer echoes occur when the bubble is aligned exactly along the axis of the ultrasound beam. (c) Sonogram shows multiple bubble clusters (arrows) that have wider, less periodic multiple echoes with lengths that depend on the size and alignment of the bubbles with respect to the ultrasound beam axis. (d) Physical explanation. A simple triad of bubbles (T) produces the periodic echoes (E) of the comet-tail artifact. Multiple clusters produce an overlying series of similar echo patterns, which eventually merge into a continuous tail (C).
|
|
View larger version (80K):
[in this window]
[in a new window]
|
Figure 9c. Bubble clusters. (a) Photograph of a small bubble cluster (arrows). (b) Sonogram shows a small bubble cluster that gives rise to a tapered distal artifact (arrows) with periodic echogenicity. Longer echoes occur when the bubble is aligned exactly along the axis of the ultrasound beam. (c) Sonogram shows multiple bubble clusters (arrows) that have wider, less periodic multiple echoes with lengths that depend on the size and alignment of the bubbles with respect to the ultrasound beam axis. (d) Physical explanation. A simple triad of bubbles (T) produces the periodic echoes (E) of the comet-tail artifact. Multiple clusters produce an overlying series of similar echo patterns, which eventually merge into a continuous tail (C).
|
|
View larger version (57K):
[in this window]
[in a new window]
|
Figure 9d. Bubble clusters. (a) Photograph of a small bubble cluster (arrows). (b) Sonogram shows a small bubble cluster that gives rise to a tapered distal artifact (arrows) with periodic echogenicity. Longer echoes occur when the bubble is aligned exactly along the axis of the ultrasound beam. (c) Sonogram shows multiple bubble clusters (arrows) that have wider, less periodic multiple echoes with lengths that depend on the size and alignment of the bubbles with respect to the ultrasound beam axis. (d) Physical explanation. A simple triad of bubbles (T) produces the periodic echoes (E) of the comet-tail artifact. Multiple clusters produce an overlying series of similar echo patterns, which eventually merge into a continuous tail (C).
|
|
A larger cluster of bubbles in contact with the ex vivo gut wall produced wider, less periodic multiple echoes with lengths that depended on the size and alignment of the bubbles with respect to the ultrasound beam axis (Fig 9c).
Physical analysis.—Surface tension causes bubbles that are in contact with each other to share surfaces. The periodic appearance of this echo is consistent with the explanation proposed by Avruch and Cooperberg (4), who suggest that a bugle-shaped space formed by the tetrahedral association of three bubbles in contact with each other acts as an acoustic resonator. The sound is transmitted back and forth along this space, and thus echoes return to the transducer at periodic intervals. The tapered waist of the resultant banded tail of echoes is probably associated with the elevation focus of the transducer (Fig 9d).
The appearance of a large cluster of bubbles can be explained by the superimposition of the multiple triads of clustered bubbles described above. The resultant artifactual streak is smeared axially by the different dimensions of the various pathways over which the multiple echoes take place (Fig 9d).
Dirty Shadowing
US appearance.—Multiple layers of bubbles in contact with the gastric wall (Fig 10a) produced multiple bright echo tails with different periods of occurrence and different strengths; these tails formed the familiar dirty shadowing that is typically seen when normal, gas-filled gut is encountered in the scanning plane (Fig 10b).
View larger version (133K):
[in this window]
[in a new window]
|
Figure 10a. Gas foam. (a) Photograph shows multiple bubbles (arrows) adhered to each other and to the mucosal wall. (b) Sonogram shows an extended layer of larger bubbles that produces a bright field of echoes, with straight margins filling the entire field of view and causing dirty shadowing (arrows). (c) Physical explanation. Scattering from many interfaces that are smaller than the wavelength of the ultrasound beam causes the reverberations and dirty shadowing, which are the classic appearances of foaming gut gas.
|
|
View larger version (171K):
[in this window]
[in a new window]
|
Figure 10b. Gas foam. (a) Photograph shows multiple bubbles (arrows) adhered to each other and to the mucosal wall. (b) Sonogram shows an extended layer of larger bubbles that produces a bright field of echoes, with straight margins filling the entire field of view and causing dirty shadowing (arrows). (c) Physical explanation. Scattering from many interfaces that are smaller than the wavelength of the ultrasound beam causes the reverberations and dirty shadowing, which are the classic appearances of foaming gut gas.
|
|
View larger version (41K):
[in this window]
[in a new window]
|
Figure 10c. Gas foam. (a) Photograph shows multiple bubbles (arrows) adhered to each other and to the mucosal wall. (b) Sonogram shows an extended layer of larger bubbles that produces a bright field of echoes, with straight margins filling the entire field of view and causing dirty shadowing (arrows). (c) Physical explanation. Scattering from many interfaces that are smaller than the wavelength of the ultrasound beam causes the reverberations and dirty shadowing, which are the classic appearances of foaming gut gas.
|
|
Physical analysis.—The quantity of gas at this point is of sufficient volume and lateral extent to prevent through transmission of the ultrasound beam. Although the gas gives a strong reflection, it is not specular because the gut-gas interface is complex, with multiple angles of incidence in the foam of bubbles. The comet tail and reverberation artifacts are superimposed on each other as before and thus produce a chaotic collection of bright streaks parallel to the axis of the ultrasound beam, with a persistent high-intensity echo (Fig 10c).
Pseudo–Thickened Gut
US appearance.—A large contiguous pocket of gas without a bubble structure in the ex vivo model produced a distinct class of artifacts in which multiple alternating hypoechoic and hyperechoic layers were seen adjacent to the gastric wall. With careful scrutiny, we recognized that the layers in the artifacts mimicked the layers of the gut wall and thus gave the false appearance of thickened gut (Fig 11a). The artifact could be composed of a single duplicated wall layer or of multiple layers. Increased enhancement of layer one was noted each time this layer appeared. Rotating the transducer about its vertical axis in the scanning plane enabled visualization of a series of identical images. These observations suggest that the pseudo–thickened gut artifact is the familiar mirror image artifact that is created at the interface between the wall and a pocket of gas.
View larger version (87K):
[in this window]
[in a new window]
|
Figure 11a. A pocket of gas. (a) Sonogram shows the layers of the real gut wall (between the arrows) repeated distal to the gas collection and causing the false appearance of thickened gut on the long axis. (b, c) Simulation of the pseudo–thickened gut artifact in a phantom. (b) Photograph of a pocket of gas (arrows) in the phantom. (c) Sonograms show the phantom without (left) and with (right) the pocket of gas. A1, A2, A3, and A4 are the artifactual components of the right image. (d) The explanation for thickened gut is the mirror image artifact. Successive reflections reach the transducer at later times and are therefore artifactually duplicated distal to their true location.
|
|
View larger version (123K):
[in this window]
[in a new window]
|
Figure 11b. A pocket of gas. (a) Sonogram shows the layers of the real gut wall (between the arrows) repeated distal to the gas collection and causing the false appearance of thickened gut on the long axis. (b, c) Simulation of the pseudo–thickened gut artifact in a phantom. (b) Photograph of a pocket of gas (arrows) in the phantom. (c) Sonograms show the phantom without (left) and with (right) the pocket of gas. A1, A2, A3, and A4 are the artifactual components of the right image. (d) The explanation for thickened gut is the mirror image artifact. Successive reflections reach the transducer at later times and are therefore artifactually duplicated distal to their true location.
|
|
View larger version (122K):
[in this window]
[in a new window]
|
Figure 11c. A pocket of gas. (a) Sonogram shows the layers of the real gut wall (between the arrows) repeated distal to the gas collection and causing the false appearance of thickened gut on the long axis. (b, c) Simulation of the pseudo–thickened gut artifact in a phantom. (b) Photograph of a pocket of gas (arrows) in the phantom. (c) Sonograms show the phantom without (left) and with (right) the pocket of gas. A1, A2, A3, and A4 are the artifactual components of the right image. (d) The explanation for thickened gut is the mirror image artifact. Successive reflections reach the transducer at later times and are therefore artifactually duplicated distal to their true location.
|
|
View larger version (21K):
[in this window]
[in a new window]
|
Figure 11d. A pocket of gas. (a) Sonogram shows the layers of the real gut wall (between the arrows) repeated distal to the gas collection and causing the false appearance of thickened gut on the long axis. (b, c) Simulation of the pseudo–thickened gut artifact in a phantom. (b) Photograph of a pocket of gas (arrows) in the phantom. (c) Sonograms show the phantom without (left) and with (right) the pocket of gas. A1, A2, A3, and A4 are the artifactual components of the right image. (d) The explanation for thickened gut is the mirror image artifact. Successive reflections reach the transducer at later times and are therefore artifactually duplicated distal to their true location.
|
|
By performing an identical procedure in the tissue-equivalent phantom with a comparable pocket of gas beneath it (Fig 11b), we demonstrated clear evidence of multiple, equally spaced mirror images (Fig 11c) of the phantom itself. We refer to this artifact as pseudo–thickened gut.
Physical analysis.—With the highly reflecting gas present at the wall-lumen interface, the echo of the multiple wall layers arrives at the transducer through two paths. The first is a direct path, through which a correctly registered image is created. The second path is created by the orthogonal reflection at the gas interface, which causes a later echo, with the layers reversed in axial order; this echo is a reflection. This is an example of the mirror image artifact (Fig 11d).
|
DISCUSSION
|
|---|
It can be seen from our experiments that the US appearance of intraluminal gas is determined by three factors: gas location, gas volume, and bubble structure. A bubble that is small relative to the wavelength of the ultrasound beam (about 0.7 mm) will act as a point scatterer and not create the mirror image artifact. A group of bubbles that are slightly larger will create a distal echo that occurs periodically according to the number and orientation of the bubbles. The pattern of multiple reflections from such a cluster of bubbles will taper as the reflections approach the focus of the transducer and finally diverge again, along with the beam, in the far field of the transducer. A larger volume of gas that is composed of many foaming bubbles will produce multiple reflections at varying times and incident angles and thereby result in the dirty shadow (Fig 10b) that obscures distal structures. In contrast, a contiguous pocket of gas, that is, one large bubble that is many wavelengths in diameter, creates the mirror image artifact of pseudo–thickened gut.
Pseudo–pneumatosis intestinalis is a misregistration artifact that is caused by elevation averaging and is produced when a bright focus—that is, a gas bubble—is present outside the scanning plane but within the elevation thickness of the ultrasound beam. Changing the angle of insonation may increase or decrease the artifact, depending on the relationship between the gas bubble and the scanning plane.
Artifacts at abdominal US are widely recognized to be useful (5). For example, identification of the shadowing distal to gallstones and of the ringdown artifact from cholesterol crystals in adenomyomatosis is requisite to correct diagnosis. Artifacts from gas in the abdomen, however, are most commonly perceived as an obstacle to US imaging. The potential for intraluminal gas to mimic a pathologic entity—that is, for pseudo–pneumatosis intestinalis or pseudo–thickened gut to occur—is not well known. Understanding the existence and origin of such artifacts should improve the clinical usefulness of abdominal US.
True pneumatosis intestinalis (Fig 1) is most commonly diagnosed with US in patients with acute abdominal pain. A clinical history that suggests the potential for vascular thrombosis or ischemia may or may not exist. The process is most often diffuse; it affects the entire gut wall in a circumferential pattern as well as a long segment of gut or several adjacent loops. Ascites, gas within the portal veins of the liver, and luminal distension of the gut are common associations. The US-based diagnosis of pneumatosis intestinalis is a difficult one and requires documentation of intramural gas. Descriptions are limited to isolated case reports (6–8). The results of our study showed that small quantities of intraluminal gas can artifactually appear to lie within the nondependent gut wall. In clinical terms, pseudo–pneumatosis intestinalis is more likely to be seen with gut wall thickening. Also, in areas of the gut that have prominent mucosal folds in which gas bubbles may be trapped, bubbles may falsely appear to be intramural (Fig 2). Therefore, although US may be sensitive in the diagnosis of pneumatosis intestinalis, artifact-based false-positive diagnoses may be made.
Differentiating pseudo–pneumatosis intestinalis from a true pathologic condition at US may be difficult. The artifactual phenomenon occurs only on the nondependent wall and usually involves the superficial wall layers in contrast to true pneumatosis intestinalis, which is typically circumferential and often has a submucosal or subserosal location. Therefore, gas within the dependent wall detected with US cannot be falsely mimicked by intraluminal gas (Fig 12). In addition, with a change in the position of the patient, the bubbles causing the pseudo–pneumatosis intestinalis artifact will be displaced and thereby will shift in location; the result is a resolution of the artifact. The use of a higher frequency transducer will improve the resolution and also may aid in distinguishing between a true pathologic condition and the illusion of disease. Angulation of the transducer also may cause the apparent location of the bubble to change.
View larger version (136K):
[in this window]
[in a new window]
|
Figure 12. True pneumatosis intestinalis with thickening of the gut wall in an 85-year-old man. Cross-sectional US image of the gut loop shows gas as bright echogenic foci (arrows) in the dependent wall, which unequivocally indicate pneumatosis intestinalis.
|
|
In clinical terms, peristalsis in the gut creates agitation of luminal contents. Just as small-bowel peristalsis is a frequent and normal occurrence, both dirty shadowing and the transient appearance of bubble echoes with tapering periodicity are common observations on sonograms obtained in healthy subjects. In pathologic conditions associated with hyperperistaltic gut, these artifacts will be exacerbated. In contrast, in clinical circumstances where the gut is quiet, a larger volume of gas without the characteristics of a bubble can accumulate. This is most pronounced when the patient is supine and motionless for an extended period, such as may occur in a prostrate patient in the intensive care unit. Such circumstances can allow the formation of a large and quiet reflecting surface of gas within the gut lumen that behaves as a mirror behind the gut wall, with the result being pseudo–thickened gut.
Other reported examples in which the mirror image reflection artifact imitates a pathologic entity have been in intimal flaps caused by a reflection at the aorta-lung interface that were detected at transesophageal echocardiography (9), apparent multiple polyps seen on endoscopic gastric US images (10), duplicated carotid or subclavian arteries at color Doppler imaging (11), reflection artifacts in the right hemidiaphragm (12) or apex of the lung (13), and misregistered intraabdominal abscesses (14). Sonographers are also familiar with the manifestation of multiple needle tips seen at US-guided biopsy.
The false appearance of thickened gut due to the mirror image artifact, which is related to a quiet pool of intraluminal gas, has been observed in numerous patients (Fig 4) at our institution and can, at times, present a diagnostic dilemma. In clinical practice, as in our study, one can demonstrate the artifactual nature of pseudo–thickened gut by reproducing the identical image in orthogonal planes. In contrast, if one is looking at a longitudinal image of true thickened gut, turning the transducer 90° will enable visualization of the thickened loop as a circle on the cross-sectional image. Further confirmation can be obtained by changing the position of the patient, which will alter the location of the gas pocket and hence the location of the artifact.
|
Footnotes
|
|---|
Address reprint requests to S.R.W.
Author contributions: Guarantors of integrity of entire study, S.R.W., P.N.B.; study concepts and design, S.R.W., P.N.B.; definition of intellectual content, S.R.W., P.N.B.; literature research, D.M.; clinical studies, S.R.W.; experimental studies, S.R.W., D.M., L.M.W., D.H.S.; data acquisition, S.R.W., D.M., L.M.W.; data analysis, S.R.W., D.M., L.M.W., P.N.B.; manuscript preparation, editing, and review, S.R.W., P.N.B.
Received July 21, 1997;
revision requested April 28, 1998; revision received July 1, 1998;
accepted August 21, 1998.
|
References
|
|---|
-
Kimmey MB, Martin RW, Haggitt RC, Wang KY, Franklin DW, Silverstein FE. Histologic correlates of gastrointestinal ultrasound images. Gastroenterology 1989; 96:433-441.[Medline]
-
Bolondi L, Casanova P, Santi V, Caletti F, Barbara L, Labo G. The sonographic appearance of the normal gastric wall: an in vitro ultrasound study. J Ultrasound Med 1986; 12:991-998.
-
Ophir J, Parker KJ. Contrast agents in diagnostic ultrasound. Ultrasound Med Biol 1989; 15:319-333.[Medline]
-
Avruch L, Cooperberg PL. The ring-down artifact. J Ultrasound Med 1985; 4:21-28.[Abstract]
-
Kremkau F, Taylor K. Artifacts in ultrasound imaging. J Ultrasound Med 1986; 5:227-237.[Abstract]
-
Vernacchia FS, Jeffrey RB, Laing FC, Wing VW. Sonographic recognition of pneumatosis intestinalis. AJR 1985; 145:51-52.[Free Full Text]
-
Sigel B, Machi J, Ramos JR, Serota AI, Robertson AL, Jr. Ultrasonic features of pneumatosis intestinalis. JCU 1985; 13:675-678.
-
Vijayaraghavan SB. Sonographic features of pneumatosis of the small bowel. JCU 1990; 18:579-581.
-
Appelby AF, Walker PF, Yeoh JK, Bonitatibus A, Yoganathan AP, Martin RP. Clinical significance and origin of artifacts in transesophageal echocardiography of the thoracic aorta. J Am Coll Cardiol 1993; 21:754-760.[Abstract]
-
Grech P. Mirror-image artifact with endoscopic ultrasonography and reappraisal of the fluid-air interface. Gastrointest Endosc 1993; 39:700-703.[Medline]
-
Middleton WD, Melson GL. The carotid ghost: a color Doppler ultrasound duplication artifact. J Ultrasound Med 1990; 9:487-493.[Abstract]
-
Gardener FJ, Clark RN, Kozlowski R. A model of a hepatic mirror-image artifact. Med Ultrasound 1980; 4:19-21.
-
Reading CC, Charboneau WJ, Allison JW, Cooperberg PL. Color and spectral Doppler mirror-image artifact of the subclavian artery. Radiology 1990; 174:41-42.[Abstract/Free Full Text]
-
Golding RH, Li DKB, Cooperberg PL. Sonographic demonstration of the air-fluid levels in abdominal abscesses. J Ultrasound Med 1982; 1:151-155.[Abstract]
This article has been cited by other articles:
|
|
|
|
 
M. Epelman, A. Daneman, O. M. Navarro, I. Morag, A. M. Moore, J. H. Kim, R. Faingold, G. Taylor, and J. T. Gerstle
Necrotizing Enterocolitis: Review of State-of-the-Art Imaging Findings with Pathologic Correlation
RadioGraphics,
March 1, 2007;
27(2):
285 - 305.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
V. M. Brandenburg, R. D. Frank, U. Janssen, P. Wurth, J. Floege, and J. Riehl
Ultrasonographic fata morgana
Nephrol. Dial. Transplant.,
April 1, 2003;
18(4):
845 - 846.
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. S. Chan, J. P. McGahan, J. Richards, and M. Gillen
The Comet Tail Artifact in the Diagnosis of Pneumothorax * Reply
J. Ultrasound Med.,
September 1, 2002;
21(9):
1060 - 1062.
[Full Text]
[PDF]
|
|
|
TOP
Abstract
Introduction
