[Insight-users] CT Lung Segmentation: Extract Outer Contours

[frédéric salvador]

Hi Luis,


1) I will definitely have a look at this experimental
itkVotingBinaryHoleFillFloodingImageFilter you suggested.

2) I dont really need to extract the surface. I was more thinking about
trying to extract the outer surface (or the outer contour in extracted 2D
slice) and fill inside afterwards.

Besides, I had some success today to fill holes by using the
itkConnectedComponentImageFilter, and it doesnt take too much time.
Though it is not well adapted to fill holes which have contacts with the
surface, hence the necessity to use morphological filters...

Anyway, thank you very much for your fast answer and suggestions.


Best regards.

Fred



Le 20 janvier 2010 20:26, Luis Ibanez <luis.ibanez at kitware.com> a écrit :

> Hi Frederic,
>
> 1) In order to fill-up holes you may want to use
>    the filter:
>
>   itkVotingBinaryIterativeHoleFillingImageFilter
>
>    or the faster (but experimental) version in the
>    NAMIC Sandbox:
>
>
> itkVotingBinaryHoleFillFloodingImageFilter.h
>
>
> http://svn.na-mic.org/NAMICSandBox/trunk/VotingHoleFillFloodingImageFilter/Source/
>
>
>
> 2)  Were you considering surface extraction
>     as an alternative method for filling-up the
>     holes ?
>
>      or do you need to do surface extraction
>      anyways ?
>
>
>
>
>    Regards,
>
>
>            Luis
>
>
> ---------------------------------------------
> 2010/1/19 frédéric salvador <frederic.salvador at gmail.com>:
> > Dear all,
> >
> >
> > So as to segment lungs in CT volume, I implemented the following steps:
> >    - low-pass smoothing,
> >    - binary thresolding,
> >    - connected component analysis,
> >    - (eventually some objects relabelling).
> >
> > So now I get a good binary mask of the lung volume.
> >
> > Though, there are numerous veins and arteries in the lungs volume
> (appearing
> > in black in my binary mask) that I would like to fill.
> >
> > I tried a rolling-ball operator consisting of a morphological closing
> > followed by a hole filling filter. But as my image is 512*512*605 it
> takes
> > too long computation time for my application. Besides, the kernel should
> be
> > pretty big so as to fill big holes.
> >
> > So here comes my question:
> > Is it possible:
> >    - to extract the outer surface, by using marching cubes for instance,
> and
> > fill inside the surface?
> > OR
> >    - to extract and operate on 2D slice to extract the outer contours, by
> > using some itk or vtk contour filter, and of course fill inside the outer
> > contours?
> >
> >
> > I have been struggling a few hours on this, so your help would be highly
> > appreciated.
> >
> > Thanks a lot in advance for your answer.
> >
> > Fred
> >
> > _____________________________________
> > 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.html
> >
> > 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 --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20100120/8a50e5fc/attachment.htm>