HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Beaulieu, C. F. Articles by Napel, S. Search for Related Content PubMed PubMed Citation Articles by Beaulieu, C. F. Articles by Napel, S. Related Collections Related Article

Display Modes for CT Colonography¹

Part II. Blinded Comparison of Axial CT and Virtual Endoscopic and Panoramic Endoscopic Volume-rendered Studies

¹ From the Department of Radiology, Rm S-056, 300 Pasteur Dr, Stanford University School of Medicine, Stanford, CA 94305. Received July 20, 1998; revision requested September 24; revision received November 18; accepted January 11, 1999. Supported in part by National Institutes of Health grants 1R01 CA72023, 1P41 RR09784-01, LM 07033; the Society for Computed Body Tomography and Magnetic Resonance; the Packard Foundation; the Lucas Foundation; and the Phil N. Allen Trust. C.F.B. is a 1997 RSNA Scholar. C.K. is a 1998 GENDEX/RSNA Medical Student/Scholar Assistant. Address reprint requests to C.F.B. (e-mail: cfb@s-word.stanford.edu).

View larger version (27K): [in a new window]   Figure 1. Schematic of virtual endoscopy. The colon is represented by a cylinder, and the position of the virtual camera is shown by the eye. The straight dashed line shows the preplanned path through the center of the colon. By orienting the camera parallel to the path, the part of the colon depicted by the shaded area is visualized. Reverse viewing is achieved by means of 180° rotation of the camera.

View larger version (15K): [in a new window]   Figure 2. Two-dimensional schematic of panoramic endoscopy. A cross-section through the colon is represented by the continuous inner circle, and the path (•, P) is in the center of this circle. The virtual camera, indicated by the eyes, is oriented perpendicular to the path. The virtual camera captures one-sixth of the circumference of the colon at each position (dashed lines, 1–6), with a 60° camera field of view. An image panel for that camera position and orientation is generated, as indicated by the interrupted outer circle (1–6). Contiguous panels displaying the colonic circumference are generated by rotating the camera in 60° increments around the path (arrows). Once generated, the six image panels are displayed side-by-side as an elongated, flattened view of the inner colonic lining. Subsequent panoramic endoscopic views are generated by moving along the path in 3-mm increments and repeating panoramic rendering.

View larger version (19K): [in a new window]   Figure 3. Schematic shows blind areas at virtual endoscopy. This longitudinal cross section of the colon contains normal haustral folds, with diagrams for forward- and reverse-viewing virtual endoscopic cameras with a 60° camera field of view. Blind, or nonvisible, areas occur for both viewing directions when the camera is oriented parallel to the path (thin dashed line), because the outer margins of the camera's visual cone (solid lines for forward viewing, thick dashed lines for reverse viewing) do not capture areas obscured by haustral folds. Blind areas for forward viewing are indicated by vertical stripes; those for reverse viewing, by horizontal stripes. While bidirectional virtual endoscopic movies help reduce the amount of blind area, a polyp such as that shown at the bottom of the diagram may not be visible, because the virtual camera never "captures" the lesion in an image frame. The exact amount of the colonic surface that cannot be visualized is a function of a complex relationship between camera field of view, the convention chosen for the viewing direction (instantaneously tangent to the path or looking toward the furthest visible path point), the amount of gross colonic curvature, the diameter of the colon, the spacing between haustral folds, and the degree to which haustra project into the lumen.

View larger version (128K): [in a new window]   Figure 4a. Axial CT display mode. (a) Magnified view shows a 3.5-mm digitally synthesized polyp (arrows) inserted along a haustral fold. (b) Magnified view shows a 10-mm digitally synthesized polyp (arrows) inserted along the colonic wall. The radiologist observers viewed full-scan field-of-view studies, on which the lesions were less conspicuous than they are on these images. A total of 355 axial sections were reviewed in stack, or cine, mode at a graphics workstation. Note that the compositing process with the base data did not create artifacts or variations in image noise that could present additional, unrealistic stimuli to the observers.

View larger version (145K): [in a new window]   Figure 4b. Axial CT display mode. (a) Magnified view shows a 3.5-mm digitally synthesized polyp (arrows) inserted along a haustral fold. (b) Magnified view shows a 10-mm digitally synthesized polyp (arrows) inserted along the colonic wall. The radiologist observers viewed full-scan field-of-view studies, on which the lesions were less conspicuous than they are on these images. A total of 355 axial sections were reviewed in stack, or cine, mode at a graphics workstation. Note that the compositing process with the base data did not create artifacts or variations in image noise that could present additional, unrealistic stimuli to the observers.

View larger version (128K): [in a new window]   Figure 5a. Virtual endoscopy display mode. Volume-rendered virtual endoscopic images of (a) a 3.5-mm polyp (arrow) and (b) a 10-mm polyp (arrows). The observers viewed animations consisting of approximately 400 image frames, of which a and b are examples, for both forward- and reverse-viewing endoscopic movies, and results obtained with both viewing directions were combined to determine the detection sensitivity.

View larger version (116K): [in a new window]   Figure 5b. Virtual endoscopy display mode. Volume-rendered virtual endoscopic images of (a) a 3.5-mm polyp (arrow) and (b) a 10-mm polyp (arrows). The observers viewed animations consisting of approximately 400 image frames, of which a and b are examples, for both forward- and reverse-viewing endoscopic movies, and results obtained with both viewing directions were combined to determine the detection sensitivity.

View larger version (66K): [in a new window]   Figure 6a. Panoramic endoscopy display mode. (a) Top: Elongated panoramic display. Bottom: Magnified view of the fifth panel from the panoramic display. A 3.5-mm synthetic polyp (arrowhead) can be seen in the fifth panel of the panoramic display and in the magnified view. The six panels represent the 360° display of the local colonic wall. Portions of a normal haustral fold (arrows) are seen in the first and sixth panels of the panoramic display. (b) Top: Elongated panoramic display. Bottom: Magnified view of the second panel from the panoramic display. A 10-mm synthetic polyp (arrowhead) is located in the second panel of the panoramic display. The observers viewed a total of 400 sequential panoramic frames for each experimental data set.

View larger version (64K): [in a new window]   Figure 6b. Panoramic endoscopy display mode. (a) Top: Elongated panoramic display. Bottom: Magnified view of the fifth panel from the panoramic display. A 3.5-mm synthetic polyp (arrowhead) can be seen in the fifth panel of the panoramic display and in the magnified view. The six panels represent the 360° display of the local colonic wall. Portions of a normal haustral fold (arrows) are seen in the first and sixth panels of the panoramic display. (b) Top: Elongated panoramic display. Bottom: Magnified view of the second panel from the panoramic display. A 10-mm synthetic polyp (arrowhead) is located in the second panel of the panoramic display. The observers viewed a total of 400 sequential panoramic frames for each experimental data set.

View larger version (21K): [in a new window]   Figure 7a. Bar graphs show detection sensitivity as a function of polyp size and display mode for (a) reader 1 and (b) reader 2. A total of 10 polyps of each size were potentially detectable in the four experimental data sets. Note that for both readers, detection increased as polyp size increased. There were significant interobserver differences in detection with the axial CT mode.

View larger version (18K): [in a new window]   Figure 7b. Bar graphs show detection sensitivity as a function of polyp size and display mode for (a) reader 1 and (b) reader 2. A total of 10 polyps of each size were potentially detectable in the four experimental data sets. Note that for both readers, detection increased as polyp size increased. There were significant interobserver differences in detection with the axial CT mode.

View larger version (22K): [in a new window]   Figure 8a. Bar graphs show detection sensitivity as a function of polyp size after correction for lesion visibility for (a) reader 1 and (b) reader 2. The number of potentially detectable polyps is given in Table 2. Because inserted but invisible lesions on renderings from a particular display mode were excluded from this analysis, detection sensitivity was higher than that for all inserted lesions (see Fig 7). Trends toward a higher degree of detection for smaller lesions are present, especially for the 3D modes.

View larger version (17K): [in a new window]   Figure 8b. Bar graphs show detection sensitivity as a function of polyp size after correction for lesion visibility for (a) reader 1 and (b) reader 2. The number of potentially detectable polyps is given in Table 2. Because inserted but invisible lesions on renderings from a particular display mode were excluded from this analysis, detection sensitivity was higher than that for all inserted lesions (see Fig 7). Trends toward a higher degree of detection for smaller lesions are present, especially for the 3D modes.