DOI: 10.1148/rg.266055199
RadioGraphics 2006;26:1669-1686
© RSNA, 2006
Three-dimensional Virtual Dissection at CT Colonography: Unraveling the Colon to Search for Lesions1
Alvin C. Silva, MD,
Clinton V. Wellnitz, MD and
Amy K. Hara, MD
1 From the Department of Diagnostic Radiology, Mayo Clinic, 13400 E Shea Blvd, Scottsdale, AZ 85259. Presented as an education exhibit at the 2004 RSNA Annual Meeting. Received November 10, 2005; revision requested February 6, 2006, and received May 10; accepted June 28. A.K.H. has a licensure agreement with GE Healthcare for the CT colonography software described in this article. All other authors have no financial relationships to disclose.
Address correspondence to A.C.S. (e-mail: silva.alvin{at}mayo.edu).
 |
Abstract
|
|---|
Computed tomographic (CT) colonography is a promising noninvasive examination for colorectal cancer screening; however, the optimal interpretation strategy remains undecided. 
Virtual dissection is an innovative technique whereby the three-dimensional (3D) model of the colon is virtually unrolled, sliced open, and displayed as a flat 3D rendering of the mucosal surface, similar to a gross pathologic specimen. This technique has the potential to reduce evaluation time by providing a more rapid 3D image assessment than is possible with an antegrade and retrograde 3D endoluminal fly-through. It may also ultimately improve accuracy by reducing blind spots present with 3D endoluminal displays and by reducing reader fatigue. A disadvantage of virtual dissection is the potential for distortion of colonic lesions and normal anatomy. To avoid potential pitfalls in image interpretation, the radiologist must be familiar with the unique appearance of the normal colon anatomy and of various pathologic findings when using virtual dissection with two-dimensional axial and 3D endoluminal CT colonographic image data sets.
© RSNA, 2006
|
Introduction
|
|---|
Current computed tomographic (CT) colonography protocols use submillimeter detector collimation, resulting in more than a thousand images for a single examination. Various image display techniques are used to interpret these large data sets. Image interpretation may include a two-dimensional (2D) axial review (1), 2D multiplanar reformation review (2,3), primary 2D review with three-dimensional (3D) comparison for problem solving (4,5), primary 3D review with 2D comparison for problem solving (6,7), computer-aided diagnosis (8), use of nonradiologists as second readers (9), and use of various 3D display options (10,11), including virtual dissection (12,13) (Fig 1). In comparison with other 3D techniques, virtual dissection may be more efficient because only a single review is performed, compared with the bidirectional review recommended for a 3D endoluminal display. Virtual dissection also allows display of the entire mucosal surface, unlike 3D endoluminal display, which may result in blind spots (10).
View larger version (44K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1a. Virtual dissection schema. (a) The virtual dissection software slices the colon open and unfolds it longitudinally by reconstructing the axial CT source image data from the perspective of a virtual camera with an orientation perpendicular to the midline of the colonic tract (T). (b) A 360° view of the inner colonic surface is presented as a flattened 3D panel with a few degrees of overlap at the edges (arrows).
|
|
View larger version (51K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1b. Virtual dissection schema. (a) The virtual dissection software slices the colon open and unfolds it longitudinally by reconstructing the axial CT source image data from the perspective of a virtual camera with an orientation perpendicular to the midline of the colonic tract (T). (b) A 360° view of the inner colonic surface is presented as a flattened 3D panel with a few degrees of overlap at the edges (arrows).
|
|
Because virtual dissection causes distortion in the morphologic depiction of the colon, familiarity with the gamut of appearances related to this technique is essential for correct interpretation. We therefore describe the appearance of the normal colon in contrast with that of common pathologic lesions, not only as depicted on virtual dissection images but also as shown on 2D axial and 3D endoluminal CT colonographic views, which are likely to be more familiar to readers. Potential imaging pitfalls (eg, incomplete distention of the colon, retained stool and fluid, and unusual artifacts) also are reviewed with regard to their appearance at virtual dissection.
|
Previous Studies of Virtual Dissection
|
|---|
Numerous rendering algorithms have been investigated to alleviate the distortion that inevitably results from the virtual straightening and flattening of curved colonic sections (14–20). One proposed method, planar virtual pathology, used isometric volume rendering, which did not require a midline colon trace (19). Proponents have concluded that this method, compared with standard 2D axial or 3D endoluminal review, has the potential for improved detection of lesions with specific morphologic characteristics (19). The image data processing method, however, requires the use of customized proprietary software, and the navigation of the reconstructed images is reportedly time consuming (19).
In a recent study in which virtual dissection review was compared with 2D axial review with endoscopic correlation, the overall time needed for analysis of the virtual dissection images was significantly greater than that for 2D images (12). However, the difference was mainly attributable to prolonged rendering times with the use of an older software version. As a result, several colonic segments were not evaluated because of reconstruction failures. With only a single monitor, this software version required eight different longitudinal panels for review, each of which displayed only a 45° section of the 360° endoluminal surface. Despite such limitations, no significant difference in polyp detection rates was found, and the average interpretation time for virtual dissection was in fact less than that for 2D review (20.9 vs 29.2 minutes) (12).
To our knowledge, these three display techniques have been compared, with endoscopic correlation, in only one other study (13). In that study, three separate reading sessions were performed in isolation: one for 2D axial images, one for 3D endoluminal views, and one for virtual dissection images. Although the authors used an older version of the software, they found that virtual dissection had the highest sensitivity and the highest positive and negative predictive values. Recently, Johnson et al (20) evaluated virtual dissection in a phantom and in a small patient sample with endoscopic correlation. All 144 polyps in the phantom and all 20 clinically proved polyps in the patients were visible in retrospect. These results indicate a high potential for improved polyp detection with this technique.
|
Colon Preparation for Virtual Dissection
|
|---|
As with images from colonoscopy and barium enema examinations, accurate interpretation of CT colonographic images requires optimal colon preparation. The standard regimen includes the observance of specific dietary restrictions before the procedure and the use of cathartics or laxatives, typically polyethylene glycol or sodium phosphate preparations. Optimal colonic distention also is necessary and can be achieved either with room air administered via a rectal tube by using a hand pump or with an automated carbon dioxide insufflator. CT colonography is performed on a multi–detector row scanner with four to 64 rows by using a low-dose technique and a single breath hold for each image acquisition (supine and prone). For further details about the scanning parameters, see the Table. Nonionic intravenous contrast material is not routinely used but may be administered for staging of a previously identified colorectal mass or for surveillance imaging. Various oral stool-tagging regimens are common; a barium suspension is used to tag stool, and meglumine diatrizoate (Gastrografin; Bracco Diagnostics, Princeton, NJ) is used to tag residual fluid in the colon.
|
Virtual Dissection for 3D Image Reconstruction
|
|---|
Recent improvements in software design have the potential to improve conspicuity and examination efficiency beyond the levels demonstrated in previous studies. The current version of a previously tested software program (CT Colonography 5.3.25; GE Healthcare, Milwaukee, Wis), with a better user interface, allows marked improvements in reconstruction time and path finding when implemented on a dual-monitor workstation (Advantage 4.2_04; GE Healthcare) (Fig 2). From the prone and supine 2D axial source image data, the virtual dissection software renders a 3D model of the colon that provides information similar to that obtained with a double-contrast barium enema examination (Fig 3). Using the rectum as the starting point and the cecum as the end point, the software automatically performs a midline tracing of a well-distended colon within less than a minute. After the result of midline tracing is confirmed and accepted at the workstation, extraneous extracolonic air-containing segments (eg, small bowel or stomach) are automatically removed by the software. If the colonic air column is discontinuous, a trace point can be manually placed distal to the obstruction on the 2D axial image, and the software then will continue the midline trace until the next discontinuity or the cecum.
View larger version (119K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2a. Workstation used for the reconstruction and display of virtual dissection images. Two side-by-side monitors allow the simultaneous interactive display of 2D axial, 3D endoluminal, and virtual dissection images. (a) On the first monitor, supine images are displayed in the top row, and prone images are displayed in the bottom row. (b) On the second monitor, a red line on the volume-rendered images (arrowheads at right) denotes the position and length of the colon segments shown on the virtual dissection images (left).
|
|
View larger version (103K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2b. Workstation used for the reconstruction and display of virtual dissection images. Two side-by-side monitors allow the simultaneous interactive display of 2D axial, 3D endoluminal, and virtual dissection images. (a) On the first monitor, supine images are displayed in the top row, and prone images are displayed in the bottom row. (b) On the second monitor, a red line on the volume-rendered images (arrowheads at right) denotes the position and length of the colon segments shown on the virtual dissection images (left).
|
|
View larger version (97K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3a. Normal anatomy and common features in a complete CT colonographic examination. (a) Volume-rendered image of the colon, from the rectum to the distal ileum. (b) Virtual dissection image of the same colon shows the rectal tube as an elongated horizontal structure that parallels the midline of the tract (red arrows). The adjacent section outlined in red (Fig. 6) shows the normal appearance of haustral folds in a straight colonic segment. Sections outlined in yellow (Fig. 5, Fig. 7, Fig. 8) and yellow arrows indicate haustral distortions related to the degree of colonic curvature and the relative position of the haustral folds on the virtual dissection image. The section outlined in green (Fig. 9) and the green arrowheads indicate areas of the colon that contain residual fluid. The outlined segments are examined in depth in subsequent figures (Figs 4–9, 16).
|
|
View larger version (55K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3b. Normal anatomy and common features in a complete CT colonographic examination. (a) Volume-rendered image of the colon, from the rectum to the distal ileum. (b) Virtual dissection image of the same colon shows the rectal tube as an elongated horizontal structure that parallels the midline of the tract (red arrows). The adjacent section outlined in red ((Fig. 6) shows the normal appearance of haustral folds in a straight colonic segment. Sections outlined in yellow (Fig. 5, Fig. 7, Fig. 8) and yellow arrows indicate haustral distortions related to the degree of colonic curvature and the relative position of the haustral folds on the virtual dissection image. The section outlined in green (Fig. 9) and the green arrowheads indicate areas of the colon that contain residual fluid. The outlined segments are examined in depth in subsequent figures (Figs 4–9, 16).
|
|
Virtual dissection views are created by visually unfolding or straightening the colon, sectioning it longitudinally, and flattening the lumen. Images obtained with the virtual camera are perpendicular to the midline of the colon (Fig 1). The 360° view of the inner colonic surface is displayed as a single flattened and colored 3D image, with a few degrees of overlap at the edges, which are indicated by a slightly darker color (Fig 1). Lesions that abut the cut surface of the colon on the virtual dissection image extend into the area of overlap (Figs 4, 5). If necessary, the entire colon can be displayed as three continuous strips on a single monitor (Fig 3). With the dual-monitor system (Fig 2), interactive interpretation can be performed by displaying 2D multiplanar, 3D endoluminal, and virtual dissection images simultaneously. Supine and prone images can be viewed in axial, coronal, sagittal, and oblique imaging planes, and preset and adjustable settings for window width and window level also are available. A real-time point-to-point correlation of findings can be performed between virtual dissection images and the corresponding 2D and 3D images (Figs 3–9). The length and position of the colon segment displayed on virtual dissection images are indicated in real time on the 3D volume-rendered image (Fig 2). The settings can be adjusted easily to obtain images of a longer or shorter segment of the colon.
View larger version (65K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4a. Rectal valves of Houston. (a) Virtual dissection image shows typical inferior (IV), middle (MV), and superior (SV) valves, which involve the lumen incompletely and have a variable orientation to the midline tract, and a rectal tube tip (arrowheads) that has a characteristic appearance. A pseudoaliasing effect is visible where the middle valve abuts the cut plane of the colon, in the region of overlap. (b) Three-dimensional endoluminal view shows the inferior valve (IV) and rectal tube (T).
|
|
View larger version (104K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4b. Rectal valves of Houston. (a) Virtual dissection image shows typical inferior (IV), middle (MV), and superior (SV) valves, which involve the lumen incompletely and have a variable orientation to the midline tract, and a rectal tube tip (arrowheads) that has a characteristic appearance. A pseudoaliasing effect is visible where the middle valve abuts the cut plane of the colon, in the region of overlap. (b) Three-dimensional endoluminal view shows the inferior valve (IV) and rectal tube (T).
|
|
View larger version (70K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5a. Distortions due to the virtual dissection reconstruction process. (a) On the virtual dissection image, the depiction of the haustral fold (arrows) in a curved segment of colon is distorted by focal angulation and pseudoaliasing effects, while the appearance of a typical, vertically oriented fold (arrowheads) in an adjacent straight segment of colon is unaffected. (b, c) Three-dimensional endoluminal view (b) and 2D axial image (c) from the same CT colonographic data set help clarify the nature of the findings on the virtual dissection image by showing normal haustral folds (arrows and arrowheads).
|
|
View larger version (96K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5b. Distortions due to the virtual dissection reconstruction process. (a) On the virtual dissection image, the depiction of the haustral fold (arrows) in a curved segment of colon is distorted by focal angulation and pseudoaliasing effects, while the appearance of a typical, vertically oriented fold (arrowheads) in an adjacent straight segment of colon is unaffected. (b, c) Three-dimensional endoluminal view (b) and 2D axial image (c) from the same CT colonographic data set help clarify the nature of the findings on the virtual dissection image by showing normal haustral folds (arrows and arrowheads).
|
|
View larger version (117K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5c. Distortions due to the virtual dissection reconstruction process. (a) On the virtual dissection image, the depiction of the haustral fold (arrows) in a curved segment of colon is distorted by focal angulation and pseudoaliasing effects, while the appearance of a typical, vertically oriented fold (arrowheads) in an adjacent straight segment of colon is unaffected. (b, c) Three-dimensional endoluminal view (b) and 2D axial image (c) from the same CT colonographic data set help clarify the nature of the findings on the virtual dissection image by showing normal haustral folds (arrows and arrowheads).
|
|
View larger version (108K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6a. Normal haustral folds. Virtual dissection image (a) and 3D endoluminal view (b) show haustral folds with a normal orientation, perpendicular to the long axis of the colon (arrowheads in a). A small amount of residual fluid partially obscures the folds (arrows).
|
|
View larger version (109K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6b. Normal haustral folds. Virtual dissection image (a) and 3D endoluminal view (b) show haustral folds with a normal orientation, perpendicular to the long axis of the colon (arrowheads in a). A small amount of residual fluid partially obscures the folds (arrows).
|
|
View larger version (89K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7a. Complex distortion of the appearance of the colon wall and haustral folds. A comparison of the virtual dissection image (a), 3D endoluminal view (b), and volume-rendered image (c) shows artifacts on the virtual dissection image that are caused by complex angulation of the colon (arrows). An apparent polypoid lesion (arrowhead in a and b) represents residual fecal matter, as indicated by its mobility between supine and prone 2D axial views (not shown).
|
|
View larger version (110K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7b. Complex distortion of the appearance of the colon wall and haustral folds. A comparison of the virtual dissection image (a), 3D endoluminal view (b), and volume-rendered image (c) shows artifacts on the virtual dissection image that are caused by complex angulation of the colon (arrows). An apparent polypoid lesion (arrowhead in a and b) represents residual fecal matter, as indicated by its mobility between supine and prone 2D axial views (not shown).
|
|
View larger version (116K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7c. Complex distortion of the appearance of the colon wall and haustral folds. A comparison of the virtual dissection image (a), 3D endoluminal view (b), and volume-rendered image (c) shows artifacts on the virtual dissection image that are caused by complex angulation of the colon (arrows). An apparent polypoid lesion (arrowhead in a and b) represents residual fecal matter, as indicated by its mobility between supine and prone 2D axial views (not shown).
|
|
View larger version (95K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8a. Complex distortion of the appearance of the colon wall and haustral folds. A simultaneous point-to-point correlation of the virtual dissection image (a) with the 3D endoluminal view (b) and volume-rendered image (c) allows an accurate interpretation of a complex fold in an angular segment of the colon (arrows), a feature that mimics a bizarrely shaped polyp or mass on the virtual dissection image.
|
|
View larger version (109K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8b. Complex distortion of the appearance of the colon wall and haustral folds. A simultaneous point-to-point correlation of the virtual dissection image (a) with the 3D endoluminal view (b) and volume-rendered image (c) allows an accurate interpretation of a complex fold in an angular segment of the colon (arrows), a feature that mimics a bizarrely shaped polyp or mass on the virtual dissection image.
|
|
View larger version (118K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8c. Complex distortion of the appearance of the colon wall and haustral folds. A simultaneous point-to-point correlation of the virtual dissection image (a) with the 3D endoluminal view (b) and volume-rendered image (c) allows an accurate interpretation of a complex fold in an angular segment of the colon (arrows), a feature that mimics a bizarrely shaped polyp or mass on the virtual dissection image.
|
|
View larger version (96K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9a. Residual fluid. (a) Virtual dissection image depicts a uniform horizontal band (arrows) indicative of residual fluid that parallels the midline of the tract and obscures portions of the haustral folds. (b) Correlative 2D axial image shows a dependent air-fluid level (arrow).
|
|
View larger version (105K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9b. Residual fluid. (a) Virtual dissection image depicts a uniform horizontal band (arrows) indicative of residual fluid that parallels the midline of the tract and obscures portions of the haustral folds. (b) Correlative 2D axial image shows a dependent air-fluid level (arrow).
|
|
|
Normal Anatomic Features on Virtual Dissection Images
|
|---|
In straight colonic segments, such as the descending or ascending colon, normal haustral folds usually are perpendicular to the midline and thus are depicted as vertically oriented lines on virtual dissection images (Figs 5, 6).
With curved colonic segments, the process of straightening and flattening may cause distortions of normal turns and haustral folds, resulting in a myriad of artifactual features (Figs 3, 5, 7, 8) (18,22). An interactive real-time correlation of the virtual dissection images with 2D and 3D images can help distinguish these findings from true polyps or masses. In well-distended normal colon segments, where haustral folds are effaced or absent, virtual dissection images are similarly featureless (Fig 3).
Normal rectal valves may appear irregular because of their variable course and length (Fig 4). Unlike haustral folds, the rectal tube (Figs 3, 4) and residual fluid (Figs 3, 6, 9) typically are parallel rather than perpendicular to the colonic long axis.
It generally is possible to differentiate these areas of distortion from true lesions with a predominantly longitudinal or nonlinear appearance by performing a real-time comparison of virtual dissection images with the corresponding 2D axial and 3D endoluminal images (Figs 10, 11).
View larger version (101K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10a. Polyp in the descending colon. Virtual dissection image (a), 3D endoluminal view (b), and 2D axial image (c) from supine CT colonography, and corresponding colonoscopy image (d), show a lesion (arrow) with a longitudinally distorted appearance on the virtual dissection image and with a broad-based attachment to the colon wall (arrowheads in a and b). It is essential that all longitudinally oriented findings on virtual dissection images be correlated with features on 3D endoluminal and 2D axial views, to differentiate true masses from pseudolesions. The 4-cm polyp was excised and histologically proved to be a sessile tubulovillous adenoma.
|
|
View larger version (118K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10b. Polyp in the descending colon. Virtual dissection image (a), 3D endoluminal view (b), and 2D axial image (c) from supine CT colonography, and corresponding colonoscopy image (d), show a lesion (arrow) with a longitudinally distorted appearance on the virtual dissection image and with a broad-based attachment to the colon wall (arrowheads in a and b). It is essential that all longitudinally oriented findings on virtual dissection images be correlated with features on 3D endoluminal and 2D axial views, to differentiate true masses from pseudolesions. The 4-cm polyp was excised and histologically proved to be a sessile tubulovillous adenoma.
|
|
View larger version (108K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10c. Polyp in the descending colon. Virtual dissection image (a), 3D endoluminal view (b), and 2D axial image (c) from supine CT colonography, and corresponding colonoscopy image (d), show a lesion (arrow) with a longitudinally distorted appearance on the virtual dissection image and with a broad-based attachment to the colon wall (arrowheads in a and b). It is essential that all longitudinally oriented findings on virtual dissection images be correlated with features on 3D endoluminal and 2D axial views, to differentiate true masses from pseudolesions. The 4-cm polyp was excised and histologically proved to be a sessile tubulovillous adenoma.
|
|
View larger version (163K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10d. Polyp in the descending colon. Virtual dissection image (a), 3D endoluminal view (b), and 2D axial image (c) from supine CT colonography, and corresponding colonoscopy image (d), show a lesion (arrow) with a longitudinally distorted appearance on the virtual dissection image and with a broad-based attachment to the colon wall (arrowheads in a and b). It is essential that all longitudinally oriented findings on virtual dissection images be correlated with features on 3D endoluminal and 2D axial views, to differentiate true masses from pseudolesions. The 4-cm polyp was excised and histologically proved to be a sessile tubulovillous adenoma.
|
|
View larger version (87K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11a. Annular adenocarcinoma in the sigmoid colon. Virtual dissection image (a), 3D endoluminal view (b), 3D volume-rendered image (c), and 2D coronal reformatted image (d) from supine CT colonography, and corresponding colonoscopy image (e), show a mass (arrow) that encircles much of the lumen but spares a small area (arrowhead in a and b). Left-sided adenocarcinomas have this classic apple-core or napkin-ring appearance more often than right-sided adenocarcinomas.
|
|
View larger version (119K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11b. Annular adenocarcinoma in the sigmoid colon. Virtual dissection image (a), 3D endoluminal view (b), 3D volume-rendered image (c), and 2D coronal reformatted image (d) from supine CT colonography, and corresponding colonoscopy image (e), show a mass (arrow) that encircles much of the lumen but spares a small area (arrowhead in a and b). Left-sided adenocarcinomas have this classic apple-core or napkin-ring appearance more often than right-sided adenocarcinomas.
|
|
View larger version (74K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11c. Annular adenocarcinoma in the sigmoid colon. Virtual dissection image (a), 3D endoluminal view (b), 3D volume-rendered image (c), and 2D coronal reformatted image (d) from supine CT colonography, and corresponding colonoscopy image (e), show a mass (arrow) that encircles much of the lumen but spares a small area (arrowhead in a and b). Left-sided adenocarcinomas have this classic apple-core or napkin-ring appearance more often than right-sided adenocarcinomas.
|
|
View larger version (149K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11d. Annular adenocarcinoma in the sigmoid colon. Virtual dissection image (a), 3D endoluminal view (b), 3D volume-rendered image (c), and 2D coronal reformatted image (d) from supine CT colonography, and corresponding colonoscopy image (e), show a mass (arrow) that encircles much of the lumen but spares a small area (arrowhead in a and b). Left-sided adenocarcinomas have this classic apple-core or napkin-ring appearance more often than right-sided adenocarcinomas.
|
|
View larger version (147K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11e. Annular adenocarcinoma in the sigmoid colon. Virtual dissection image (a), 3D endoluminal view (b), 3D volume-rendered image (c), and 2D coronal reformatted image (d) from supine CT colonography, and corresponding colonoscopy image (e), show a mass (arrow) that encircles much of the lumen but spares a small area (arrowhead in a and b). Left-sided adenocarcinomas have this classic apple-core or napkin-ring appearance more often than right-sided adenocarcinomas.
|
|
Johnson et al (20) described haustral distortion as occurring longitudinally on virtual dissection images. Polyp distortion on virtual dissection images was dependent on the underlying morphologic features, with sessile and flat masses appearing as flame- or pea-shaped lesions, and with pedunculated masses generally appearing as flame- or club-shaped lesions. The few bizarrely shaped lesions observed on virtual dissection images often represented a pedunculated polyp. Thus, with training, the radiologist would not find that distortion on virtual dissection images compromises polyp detection but rather that virtual dissection may allow a more rapid review because there are fewer images to interpret.
|
Pitfalls of Virtual Dissection
|
|---|
Like 3D images obtained with other CT colonographic display techniques, virtual dissection images may be affected by incomplete or suboptimal colonic distention, which is a common cause of inaccurate depiction. On virtual dissection images, a "bridge " links discontinuous colonic segments, allowing the software to continue the mid-line trace, with collapsed areas appearing as featureless gaps in the colon (Fig 12).
View larger version (45K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 12a. Incomplete distention of the colon. Virtual dissection image (a) and 3D volume-rendered image (b) from CT colonography show a featureless gap (arrows) roughly proportional in length to the colon segment that was not distended.
|
|
View larger version (130K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 12b. Incomplete distention of the colon. Virtual dissection image (a) and 3D volume-rendered image (b) from CT colonography show a featureless gap (arrows) roughly proportional in length to the colon segment that was not distended.
|
|
Because untagged residual fluid (Fig 13) may obscure lesions, it is necessary to use supine and prone positioning of the patient, fluid tagging, or both to ensure the identification of all lesions. Residual fecal matter can be identified by its mobility with a change in patient position, its heterogeneity (Fig 14), or its increased attenuation due to barium stool tagging.
View larger version (143K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13a. Pedunculated polyp obscured by residual fluid in the ascending colon. (a–d) Virtual dissection image (a), 3D endoluminal image (b), and 2D axial image (c) from prone CT colonography, and corresponding colonoscopy image (d), demonstrate a pedunculated polyp (arrow). Residual fluid is visible in a dependent position in a and c (arrowheads). (e) Two-dimensional axial image from supine CT colonography shows only the tip of the polyp (arrow), most of which is submerged in residual fluid (arrowhead). The polyp was likewise obscured by fluid on supine virtual dissection and 3D endoluminal images (not shown). The case illustrates the necessity of reviewing both prone and supine images for accurate CT colonographic interpretation.
|
|
View larger version (123K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13b. Pedunculated polyp obscured by residual fluid in the ascending colon. (a–d) Virtual dissection image (a), 3D endoluminal image (b), and 2D axial image (c) from prone CT colonography, and corresponding colonoscopy image (d), demonstrate a pedunculated polyp (arrow). Residual fluid is visible in a dependent position in a and c (arrowheads). (e) Two-dimensional axial image from supine CT colonography shows only the tip of the polyp (arrow), most of which is submerged in residual fluid (arrowhead). The polyp was likewise obscured by fluid on supine virtual dissection and 3D endoluminal images (not shown). The case illustrates the necessity of reviewing both prone and supine images for accurate CT colonographic interpretation.
|
|
View larger version (123K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13c. Pedunculated polyp obscured by residual fluid in the ascending colon. (a–d) Virtual dissection image (a), 3D endoluminal image (b), and 2D axial image (c) from prone CT colonography, and corresponding colonoscopy image (d), demonstrate a pedunculated polyp (arrow). Residual fluid is visible in a dependent position in a and c (arrowheads). (e) Two-dimensional axial image from supine CT colonography shows only the tip of the polyp (arrow), most of which is submerged in residual fluid (arrowhead). The polyp was likewise obscured by fluid on supine virtual dissection and 3D endoluminal images (not shown). The case illustrates the necessity of reviewing both prone and supine images for accurate CT colonographic interpretation.
|
|
View larger version (149K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13d. Pedunculated polyp obscured by residual fluid in the ascending colon. (a–d) Virtual dissection image (a), 3D endoluminal image (b), and 2D axial image (c) from prone CT colonography, and corresponding colonoscopy image (d), demonstrate a pedunculated polyp (arrow). Residual fluid is visible in a dependent position in a and c (arrowheads). (e) Two-dimensional axial image from supine CT colonography shows only the tip of the polyp (arrow), most of which is submerged in residual fluid (arrowhead). The polyp was likewise obscured by fluid on supine virtual dissection and 3D endoluminal images (not shown). The case illustrates the necessity of reviewing both prone and supine images for accurate CT colonographic interpretation.
|
|
View larger version (146K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13e. Pedunculated polyp obscured by residual fluid in the ascending colon. (a–d) Virtual dissection image (a), 3D endoluminal image (b), and 2D axial image (c) from prone CT colonography, and corresponding colonoscopy image (d), demonstrate a pedunculated polyp (arrow). Residual fluid is visible in a dependent position in a and c (arrowheads). (e) Two-dimensional axial image from supine CT colonography shows only the tip of the polyp (arrow), most of which is submerged in residual fluid (arrowhead). The polyp was likewise obscured by fluid on supine virtual dissection and 3D endoluminal images (not shown). The case illustrates the necessity of reviewing both prone and supine images for accurate CT colonographic interpretation.
|
|
View larger version (138K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 14a. Residual fecal matter in the transverse colon. (a, b) Virtual dissection image (a) and 3D endoluminal view (b) from CT colonography demonstrate a polypoid filling defect (arrow). (c) Axial image from CT colonography shows a heterogeneous defect with a central area of hypoattenuation (arrow) that is characteristic of stool.
|
|
View larger version (126K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 14b. Residual fecal matter in the transverse colon. (a, b) Virtual dissection image (a) and 3D endoluminal view (b) from CT colonography demonstrate a polypoid filling defect (arrow). (c) Axial image from CT colonography shows a heterogeneous defect with a central area of hypoattenuation (arrow) that is characteristic of stool.
|
|
View larger version (119K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 14c. Residual fecal matter in the transverse colon. (a, b) Virtual dissection image (a) and 3D endoluminal view (b) from CT colonography demonstrate a polypoid filling defect (arrow). (c) Axial image from CT colonography shows a heterogeneous defect with a central area of hypoattenuation (arrow) that is characteristic of stool.
|
|
The ileocecal valve has various appearances, and it may resemble a mass on virtual dissection images. It is typically necessary, therefore, to review 2D axial images for confirmation of the appearance of the ileocecal valve, particularly in the presence of lipomatous infiltration (Figs 15, 16).
View larger version (143K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 15a. Lipomatous infiltration of the ileocecal valve. Virtual dissection image (a), 3D endoluminal image (b), and 2D axial image (c) from CT colonography demonstrate a polypoid filling defect (arrow) that, in c, has attenuation characteristic of fat.
|
|
View larger version (133K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 15b. Lipomatous infiltration of the ileocecal valve. Virtual dissection image (a), 3D endoluminal image (b), and 2D axial image (c) from CT colonography demonstrate a polypoid filling defect (arrow) that, in c, has attenuation characteristic of fat.
|
|
View larger version (153K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 15c. Lipomatous infiltration of the ileocecal valve. Virtual dissection image (a), 3D endoluminal image (b), and 2D axial image (c) from CT colonography demonstrate a polypoid filling defect (arrow) that, in c, has attenuation characteristic of fat.
|
|
View larger version (49K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 16a. Ileocecal valve. Virtual dissection image (a), 3D endoluminal view (b), and 2D axial image (c) from CT colonography depict the lumen of the ileocecal valve (arrow). A tiny nearby filling defect (arrowhead in a and b), which demonstrated mobility with changes in patient positioning, represents residual fecal matter.
|
|
View larger version (104K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 16b. Ileocecal valve. Virtual dissection image (a), 3D endoluminal view (b), and 2D axial image (c) from CT colonography depict the lumen of the ileocecal valve (arrow). A tiny nearby filling defect (arrowhead in a and b), which demonstrated mobility with changes in patient positioning, represents residual fecal matter.
|
|
View larger version (129K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 16c. Ileocecal valve. Virtual dissection image (a), 3D endoluminal view (b), and 2D axial image (c) from CT colonography depict the lumen of the ileocecal valve (arrow). A tiny nearby filling defect (arrowhead in a and b), which demonstrated mobility with changes in patient positioning, represents residual fecal matter.
|
|
A postoperative appendiceal stump may mimic a polyp (Fig 17). In cases in which a postoperative stump is suspected, a comprehensive clinical history and the location of the finding at the cecal base should facilitate correct diagnosis.
View larger version (132K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 17a. Appendiceal stump. Virtual dissection image (a), 3D endoluminal view (b), and axial image (c) from CT colonography, and endoscopic image (d), show a polypoid filling defect (arrow) at the cecal base, a finding that represents a pseudolesion from a previous appendectomy. Note the characteristic triangular appearance of the haustral folds at this location.
|
|
View larger version (135K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 17b. Appendiceal stump. Virtual dissection image (a), 3D endoluminal view (b), and axial image (c) from CT colonography, and endoscopic image (d), show a polypoid filling defect (arrow) at the cecal base, a finding that represents a pseudolesion from a previous appendectomy. Note the characteristic triangular appearance of the haustral folds at this location.
|
|
View larger version (128K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 17c. Appendiceal stump. Virtual dissection image (a), 3D endoluminal view (b), and axial image (c) from CT colonography, and endoscopic image (d), show a polypoid filling defect (arrow) at the cecal base, a finding that represents a pseudolesion from a previous appendectomy. Note the characteristic triangular appearance of the haustral folds at this location.
|
|
View larger version (141K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 17d. Appendiceal stump. Virtual dissection image (a), 3D endoluminal view (b), and axial image (c) from CT colonography, and endoscopic image (d), show a polypoid filling defect (arrow) at the cecal base, a finding that represents a pseudolesion from a previous appendectomy. Note the characteristic triangular appearance of the haustral folds at this location.
|
|
Finally, artifacts caused by reconstruction errors may result in a false-positive finding of a pathologic entity (Fig 18). In addition, diverticula occasionally simulate polyps on virtual dissection images, but correlation with the 2D axial images allows easy distinction between these two entities (Fig 19).
View larger version (122K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 18a. Artifacts mimicking polyposis. (a, b) Virtual dissection image (a) and 3D endoluminal view (b) from a supine CT colonographic data set reconstructed with 4-mm collimation at 4-mm intervals show apparent polyps (arrowheads). (c, d) Virtual dissection image (c) and 3D endoluminal view (d) from a prone data set reconstructed with 1-mm collimation at 1-mm intervals show the absence of polyps with the thin-section reconstruction.
|
|
View larger version (124K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 18b. Artifacts mimicking polyposis. (a, b) Virtual dissection image (a) and 3D endoluminal view (b) from a supine CT colonographic data set reconstructed with 4-mm collimation at 4-mm intervals show apparent polyps (arrowheads). (c, d) Virtual dissection image (c) and 3D endoluminal view (d) from a prone data set reconstructed with 1-mm collimation at 1-mm intervals show the absence of polyps with the thin-section reconstruction.
|
|
View larger version (109K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 18c. Artifacts mimicking polyposis. (a, b) Virtual dissection image (a) and 3D endoluminal view (b) from a supine CT colonographic data set reconstructed with 4-mm collimation at 4-mm intervals show apparent polyps (arrowheads). (c, d) Virtual dissection image (c) and 3D endoluminal view (d) from a prone data set reconstructed with 1-mm collimation at 1-mm intervals show the absence of polyps with the thin-section reconstruction.
|
|
View larger version (110K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 18d. Artifacts mimicking polyposis. (a, b) Virtual dissection image (a) and 3D endoluminal view (b) from a supine CT colonographic data set reconstructed with 4-mm collimation at 4-mm intervals show apparent polyps (arrowheads). (c, d) Virtual dissection image (c) and 3D endoluminal view (d) from a prone data set reconstructed with 1-mm collimation at 1-mm intervals show the absence of polyps with the thin-section reconstruction.
|
|
View larger version (89K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 19a. Diverticulosis in the sigmoid colon. Virtual dissection image (a) and 3D endoluminal view (b) from supine CT colonography show diverticula (arrowheads). Frequently found adjacent to the teniae coli, diverticula often are arranged in a linear fashion that is more apparent on virtual dissection images (translucent line in a) than on other CT colonographic views.
|
|
View larger version (106K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 19b. Diverticulosis in the sigmoid colon. Virtual dissection image (a) and 3D endoluminal view (b) from supine CT colonography show diverticula (arrowheads). Frequently found adjacent to the teniae coli, diverticula often are arranged in a linear fashion that is more apparent on virtual dissection images (translucent line in a) than on other CT colonographic views.
|
|
|
Conclusions
|
|---|
Technological advances in CT colonography and the increasing public acceptance of its use have aided its evolution as a potential alternative screening examination for colorectal carcinoma. With the growing use of CT colonography, accurate and time-efficient interpretation strategies are needed to accommodate the expected increase in the number of examinations performed with this technique. In that context, virtual dissection appears promising. However, a familiarity with the altered appearance of normal and abnormal colon features at virtual dissection is critical for accurate image interpretation.
|
Footnotes
|
|---|
Abbreviations: 3D = three-dimensional, 2D = two-dimensional
|
References
|
|---|
- Bruzzi JF, Moss AC, Brennan DD, MacMathuna P, Fenlon HM. Colonic surveillance by CT colonography using axial images only. Eur Radiol 2004;14:763–767.[CrossRef][Medline]
- McFarland EG, Brink JA, Pilgram TK, et al. Spiral CT colonography: reader agreement and diagnostic performance with two- and three-dimensional image-display techniques. Radiology 2001; 218:375–383.[Abstract/Free Full Text]
- Fletcher JG, Johnson CD, Welch TJ, et al. Optimization of CT colonography technique: prospective trial in 180 patients. Radiology 2000;216: 704–711.[Abstract/Free Full Text]
- Barish MA, Soto JA, Ferrucci JT. Consensus on current clinical practice of virtual colonoscopy. AJR Am J Roentgenol 2005;184:786–792.[Abstract/Free Full Text]
- Macari M, Milano A, Lavelle M, Berman P, Megibow AJ. Comparison of time-efficient CT colonography with two- and three-dimensional colonic evaluation for detecting colorectal polyps. AJR Am J Roentgenol 2000;174:1543–1549.[Abstract/Free Full Text]
- Dachman AH, Kuniyoshi JK, Boyle CM, et al. CT colonography with three-dimensional problem solving for detection of colonic polyps. AJR Am J Roentgenol 1998;171:989–995.[Abstract/Free Full Text]
- Pickhardt PJ, Choi JR, Hwang I, Schindler WR. Nonadenomatous polyps at CT colonography: prevalence, size distribution, and detection rates. Radiology 2004;232:784–790.[Abstract/Free Full Text]
- Blake ME, Soto JA, Hayes RA, Ferrucci JT. Automated volumetry at CT colonography: a phantom study. Acad Radiol 2005;12:608–613.[CrossRef][Medline]
- Bodily KD, Fletcher JG, Engelby T, et al. Nonradiologists as second readers for intraluminal findings at CT colonography. Acad Radiol 2005;12: 67–73.[CrossRef][Medline]
- Beaulieu CF, Jeffrey RB Jr, Karadi C, Paik DS, Napel S. Display modes for CT colonography. II. Blinded comparison of axial CT and virtual endoscopic and panoramic endoscopic volume-rendered studies. Radiology 1999;212:203–212.[Abstract/Free Full Text]
- Vos FM, van Gelder RE, Serlie IW, et al. Three-dimensional display modes for CT colonography: conventional 3D virtual colonoscopy versus unfolded cube projection. Radiology 2003;228:878–885.[Abstract/Free Full Text]
- Hoppe H, Quattropani C, Spreng A, Mattich J, Netzer P, Dinkel HP. Virtual colon dissection with CT colonography compared with axial interpretation and conventional colonoscopy: preliminary results. AJR Am J Roentgenol 2004;182:1151–1158.[Abstract/Free Full Text]
- Rottgen R, Fischbach F, Plotkin M, et al. CT colonography using different reconstruction modi. Clin Imaging 2005;29:195–199.[CrossRef][Medline]
- Dave SB, Wang G, Brown BP, McFarland EG, Zhang Z, Vannier MW. Straightening the colon with curved cross sections: an approach to CT colonography. Acad Radiol 1999;6:398–410.[CrossRef][Medline]
- Wang G, McFarland EG, Brown BP, Vannier MW. GI tract unraveling with curved cross sections. IEEE Trans Med Imaging 1998;17:318–322.[CrossRef][Medline]
- Zhang Z, Wang G, Brown BP, Vannier MW. Distortion reduction for fast soft straightening of the colon. Acad Radiol 2000;7:506–515.[CrossRef][Medline]
- Zhang Z, Wang G, Brown BP, McFarland EG, Haller J, Vannier MW. Fast algorithm for soft straightening of the colon. Acad Radiol 2000;7: 142–148.[CrossRef][Medline]
- Paik DS, Beaulieu CF, Jeffrey RB Jr, Karadi CA, Napel S. Visualization modes for CT colonography using cylindrical and planar map projections. J Comput Assist Tomogr 2000;24:179–188.[CrossRef][Medline]
- Fletcher JG, Johnson CD, Reed JE, Garry J. Feasibility of planar virtual pathology: a new paradigm in volume-rendered CT colonography. J Comput Assist Tomogr 2001;25:864–869.[CrossRef][Medline]
- Johnson KT, Johnson CD, Fletcher JG, MacCarty RL, Summers RL. CT colonography using 360-degree virtual dissection: a feasibility study. AJR Am J Roentgenol 2006;186:90–95.[Abstract/Free Full Text]
- Silva AC, Hara AK, Leighton JA, Heppell JP. CT colonography with intravenous contrast material: varied appearances of colorectal carcinoma. RadioGraphics 2005;25:1321–1334.[Abstract/Free Full Text]
- Macari M, Megibow AJ. Pitfalls of using three-dimensional CT colonography with two-dimensional imaging correlation. AJR Am J Roentgenol 2001;176:137–143.[Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
D. Hock, R. Ouhadi, R. Materne, A.-S. Aouchria, I. Mancini, T. Broussaud, P. Magotteaux, and A. Nchimi
Virtual Dissection CT Colonography: Evaluation of Learning Curves and Reading Times with and without Computer-aided Detection
Radiology,
September 1, 2008;
248(3):
860 - 868.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
P. J. Pickhardt, A. D. Lee, A. J. Taylor, S. J. Michel, T. C. Winter, A. Shadid, R. J. Meiners, P. J. Chase, J. L. Hinshaw, J. G. Williams, et al.
Primary 2D Versus Primary 3D Polyp Detection at Screening CT Colonography
Am. J. Roentgenol.,
December 1, 2007;
189(6):
1451 - 1456.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
P. J. Pickhardt
Screening CT Colonography: How I Do It
Am. J. Roentgenol.,
August 1, 2007;
189(2):
290 - 298.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
