<table cellspacing="0" cellpadding="0" border="0" ><tr><td valign="top" style="font: inherit;">Hi Gib,<div><br></div><div>In addition to what Brian and Richard said, maybe using a different segmentation method</div><div><br></div><div>can improve the results altogether since you have no problems with small-diameter vessels</div><div><br></div><div>and your problem is with the vessels of large diameter, even when they would have no</div><div><br></div><div>holes. You haven't mentioned what segmentation method you use in your preprocessing</div><div><br></div><div>step, my suggestion is to incorporate a priori in the segmentation based on a initial</div><div><br></div><div>estimation of a binary edge image with its gaps filled (which would be easier than filling the</div><div><br></div><div>gaps in a binary image). I think there are two segmentation methods already implemented</div><div><br></div><div>in ITK with the
ability to incorporate a priori, geodesic active contour with shape guidance</div><div><br></div><div>and shape prior level set. Also, the Canny edge detector gives a good estimation of the</div><div><br></div><div>edges.</div><div><br></div><div>Regards,</div><div><br></div><div>Dawood</div><div><br></div><div><br></div><div>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>><<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<</div><div><br></div><div>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>><<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<</div><div><br></div><div><br></div><div><span class="Apple-style-span" style="font-family: 'Times
New Roman'; font-size: medium; "><pre>On Wed, Feb 2, 2011 at 10:28 PM, Gib Bogle <<a href="http://www.itk.org/mailman/listinfo/insight-users">g.bogle at auckland.ac.nz</a>> wrote:
><i> I'm working with a volume image that was generated by labelling the laminae
</i>><i> of blood vessels. My aim is to segment out the vasculature. There are many
</i>><i> difficulties, and the particular issue I'm addressing at the moment is
</i>><i> filling in the vessels. The intensity of the labelling of the walls is
</i>><i> variable, with patches that are indistinguishable from background. The
</i>><i> vessel diameters vary widely, from about 4 to about 60 voxels. After some
</i>><i> preprocessing I have a binary image, on which the ITK hole-filling function
</i>><i> works well with the small-diameter vessels, but the big vessels present a
</i>><i> problem, even when the walls are "watertight" (i.e. without holes).
</i>><i>
</i>><i> My best idea so far is to send probes out in all 26 directions (all
</i>><i> neighbours of a voxel) and count the number of probes that hit a wall within
</i>><i> a specified radius. It's tricky to specify both the radius and the critical
</i>><i> number of hits, without getting too many false positives (voxels outside the
</i>><i> vessels showing up as inside). (The filter is of course applied
</i>><i> iteratively.)
</i>><i>
</i>><i> I'm wondering if anyone else here has addressed a similar problem.
</i>><i>
</i>><i> Thanks
</i>><i> Gib</i></pre></span></div></td></tr></table><br>