header > abstract > paper > figures and movies > footer

Virtual Colon Unfolding

by A. Vilanova Bartroli, R. Wegenkittl, A. König, and E. Gröller
published at IEEE Visualization 2001

This page assembles some animations results of work that is part of our research topic ``Visualization of Medical Data''. The figures are provided in GIF or JPG format and the movies are MPEG1 compressed (mpg).

header < abstract > paper > figures and movies > footer

Abstract of the Paper:

The majority of virtual endoscopy techniques tries to simulate a real endoscopy. A real endoscopy does not always give the optimal information due to the physical limitations it is subject to. In this paper, we deal with the unfolding of the surface of the colon as a possible visualization technique for diagnosis and polyp detection. A new two-step technique is presented which deals with the problems of double appearance of polyps and nonuniform sampling that other colon unfolding techniques suffer from. In the first step, a distance map from a central path induces nonlinear rays for unambiguous parameterization of the surface. The second step compensates for locally varying distortions of the unfolded surface. A technique similar to magnification fields in information visualization is hereby applied. The technique produces a single view of a complete virtually dissected colon.

header < abstract < paper > figures and movies > footer

Download the paper (.ps.zip) (preferred) (~3.396KB)

Download the paper (.pdf) (lower quality images)(~2.435KB)

header < abstract < paper < figures and movies > footer

Images and Animations

Movie illustrating the virtual colon unfolding method (mpeg4 codec, .avi). (~17.528KB)

(.mpg) (~25.623KB)
[nonlinear 2D Scaling]
Portion of the movie showing the evolution of the 2D grid during the nonlinear 2D scaling (.mpg). (~3.290KB)
High resolution image of the virtual unfolded colon

[Extracted Colon]


Figures in the paper

[Fig. 1]
Figure 1 :
Outside view and virtual unfolding of a segment of a CT data set of a healthy colon with a resolution of 198x115x300.
[Fig. 2]
Figure 2 :
Illustration of the possible undersampling and double appearance of polyps due to intersections of the cross-sections in high curvature areas. The dashed cross-section line produces a double appearance of a polyp.
[Fig. 3]
Figure 3 :
Illustration of the elimination of double polyp appearance by nonlinear ray casting.
[Fig. 4]
Figure 4 :
The nonlinear rays traced from a specific path position:
a) the curved rays in areas of high curvature for two consecutive points along C(v), b) nonlinear rays traced in u direction.
[Fig. 5]
Figure 5 :
Surface obtained after nonlinear ray casting. It can be observed that the sampling of the surface is nonuniform.
[Fig. 6]
Figure 6 :
a) Unfolding of the colon surface of the data set presented in figure 5 using straight rays. Solid circles indicates double polyp appearance areas. Dashed circles indicate undersampled areas. b) unfolding of the parametric surface shown in figure 5 by mapping the parametric space to a regular grid.
[Fig. 7]
Figure 7 :
Illustration of height field unfolding:
a) cross-section of r(u,v) for a fixed value of v, b) unfolding to a regular grid, c) unfolding preserving the edge lengths of the 3D quadrilateral mesh in the 2D grid d) unfolding of the height field preserving the edge lengths of the 3D quadrilateral mesh.
[Fig. 8]
Figure 8 :
Illustration of height field unfolding in the v direction.
[Fig. 9a][Fig. 9b][Fig. 9c] Figure 9a, Figure 9b and Figure 9c:
Illustration of the nonlinear 2D scaling algorithm using the same data set than in figure 5. a) Initial Tc corresponding to a 128x171 grid. b) Tc after 960 iterations of the algorithm. c) Tc after 1687 iterations.
[Fig. 10a][Fig. 10b][Fig. 10c] Figure 10a, Figure 10b and Figure 10c:
Resampling after the nonlinear 2D scaling. a) 128x171 shaded grid using bilinear interpolation. b) the resulting grid of the nonlinear 2D scaling after resampling c) Shading of the resampled grid.
[Fig. 11]
Figure 11 :
Outside view of the segmented surface of the extracted colon CT data set with resolution 381x120x632.
[Fig. 12]
Figure 12 :
Left Virtually unfolded extracted colon with the polyps numbered according to the real dissection.
Right qualitative comparison of the virtually unfolded colon with pictures taken from the real dissection. Right bottom image corresponding to the segment of the extracted colon data set used throughout this paper as example. The orientation in which the pictures were taken does not correspond with the orientation of the virtually unfolded colon.

header < abstract < paper < figures and movies < footer

This page was last updated by Anna Vilanova on June 13, 2002.
If you have any comments, please send a message to anna@cg.tuwien.ac.at.