[Insight-users] Geodesic Active Contour giving negative levelset both inside and outside of vessel and padded boundary

[Dan Mueller]

Hi Liza,

Your initial levelset should be a signed function: inside it should be
negative, zero on the contour, and positive outside initial contour. I
have a little helper function which converts a binary image into an
initial levelset using the SignedMaurerDistanceImageFilter (make sure
you set InsideIsPositive = false).

The output of the levelset operation is also a signed function:
negative inside the contour, zero on the contour, and positive outside
the contour. To retrieve the entire object, you need to apply a binary
threshold operation on the resultant levelset function: use a
BinaryThresholdImageFilter with LowerThreshold = -1000.0,
UpperThreshold = 0.0, InsideValue = 255, and OutsideValue = 0. This
will return all the pixels inside and on the final contour.

It may be that the final levelset contour you are visualizing is only
the pixels on the contour (i.e. value = 0.0). Make sure you apply a
binary threshold so that all pixels in the range [-1000,0.0] are
included in the image.

HTH

Cheers, Dan

On 22 June 2013 00:46, liza <lizashr at outlook.com> wrote:
>
>
>
>
>
> Dear All,
>
> I am trying to segment vessel with aneurysm using Geodesic Active Contour Level Set filter. My vessel doesn't have closed ends but it runs until the end of image. I've tried putting seeds well within the vessel. I have set the stopping value in fastmarching to be twice the seed radius. Since the radius of the vessel is not constant as it has aneurysm, I am not sure how small or large the stopping value should be. If I set the radius to be less than 1 then the resulting image is looks like the object that I want to segment as the aneurysm is captured well but the arteries very tiny. If I increase the size of the seed to be greater than 1 then I do get the entire object, however, I get negative levelset values both inside of aneurysm as well as outside and the boundaries are padded to positive value. I wonder if it because I increased the radius size. I have set advection and curvature scaling to 1 and have increased propagation scaling to 50.
> Has anyone encountered such issue before?
>
> I'd greatly appreciate your thoughts on this.
>
> Thanks,
> Liza
>
>
>
>
> -----
> Liza
> --
> View this message in context: http://itk-insight-users.2283740.n2.nabble.com/Geodesic-Active-Contour-giving-negative-levelset-both-inside-and-outside-of-vessel-and-padded-bounday-tp7583356.html
> Sent from the ITK Insight Users mailing list archive at Nabble.com.
> _____________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.html
>
> Kitware offers ITK Training Courses, for more information visit:
> http://www.kitware.com/products/protraining.php
>
> Please keep messages on-topic and check the ITK FAQ at:
> http://www.itk.org/Wiki/ITK_FAQ
>
> Follow this link to subscribe/unsubscribe:
> http://www.itk.org/mailman/listinfo/insight-users
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 01 Initial level set.jpg
Type: image/jpeg
Size: 6956 bytes
Desc: not available
URL: <http://www.itk.org/pipermail/insight-users/attachments/20130622/c3e18f1b/attachment-0003.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 02 Final level set.jpg
Type: image/jpeg
Size: 17523 bytes
Desc: not available
URL: <http://www.itk.org/pipermail/insight-users/attachments/20130622/c3e18f1b/attachment-0004.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 03 Final level set.jpg
Type: image/jpeg
Size: 3862 bytes
Desc: not available
URL: <http://www.itk.org/pipermail/insight-users/attachments/20130622/c3e18f1b/attachment-0005.jpg>