[Insight-users] How to create a LabelMap or Label object

Anja Ende anja.ende at googlemail.com
Fri May 2 05:27:58 EDT 2008


Hi,

The points actually define the boundary of the irregular shape. So, I
have the algorithm that will extract all the voxels that are covered
by this shape boundary. What I basically need to do is somehow
efficiently extract this bit of the image into a label map, I guess.
So given a set of points that define the ROI (like a
BlobSpatialObject), how can I convert it to the data structure that I
can use with the statistics bit.

Cheers,
Anja

2008/5/2 Richard Beare <richard.beare at gmail.com>:
> Hi,
>
>  Do you mean that the points define the corners of a polygon region?
>  If so your best bet is to render this shape into an image and then
>  proceed to the measurement phase using the tools that have been
>  discussed in this thread. This may not sound optimal, but will
>  certainly be simplest and  actually quite fast - the translation to
>  connected component and measurement phases should only require one
>  image visit each. My suggestion is to see whether this does the  job
>  for you and worry about efficiency later. If you decide to make use of
>  the point information directly you will need to make a number of
>  decisions about how you deal with voxels along the borders etc. It
>  will probably get quite messy and time consuming.
>
>
>
>  On Fri, May 2, 2008 at 4:43 PM, Anja Ende <anja.ende at googlemail.com> wrote:
>  > Hi Xabi,
>  >
>  >  Thanks for the reply. However, is it not possible to create a label
>  >  simply based on the points. When the user delineates the object, I
>  >  usually have the connected points and it would be much more efficient
>  >  for me to use these points.
>  >
>  >  Cheers,
>  >  Anja
>  >
>  >  2008/5/1 Xabier Artaechevarria Artieda <xabiarta at unav.es>:
>  >  > Hi Anja,
>  >  >  Look at section 5.1 of the article. The easiest thing is to convert a label
>  >  > image or binary image (maybe this is your case) into a itk::LabelMap using
>  >  > itk::BinaryImageToLabelMapFilter or itk::LabelImageToLabelMapFilter.
>  >  >  Then, in your case, you will need to convert your image into a binary
>  >  > image.
>  >  >  And last, use the mentioned filter to convert it to a LabelMap.
>  >  >  Regards,
>  >  >  Xabi
>  >  >
>  >  >  --
>  >  >  Xabier Artaechevarria
>  >  >  Cancer Imaging Laboratory
>  >  >  Centre for Applied Medical Research
>  >  >  www.cima.es
>  >  >
>  >  >
>  >  >
>  >  >  Anja Ende <anja.ende at googlemail.com> ha escrito:
>  >  >
>  >  >
>  >  > >
>  >  > >
>  >  > >
>  >  > > Hi everyone,
>  >  > >
>  >  > > I had a question a few days back where I needed to perform some
>  >  > > statistics an irregular region of an image. Someone had pointed me to
>  >  > > the following submission:
>  >  > >
>  >  > >
>  >  > http://www.insight-journal.org/InsightJournalManager/view_reviews.php?back=index.php&pubid=176
>  >  > >
>  >  > > However, I am a bit confused on how to use it. So, if I understand
>  >  > > correctly, I need to create a LabelMap object.
>  >  > >
>  >  > > What I do have is the set of connected points (in image coordinates)
>  >  > > that make up my region and basically represents a closed 2D polygon or
>  >  > > surface. How can I use that information to create this LabelMap class?
>  >  > >
>  >  > > Any help would be really appreciated.
>  >  > >
>  >  > > Cheers,
>  >  > >
>  >  > > Anja
>  >  > > _______________________________________________
>  >  > > Insight-users mailing list
>  >  > > Insight-users at itk.org
>  >  > > http://www.itk.org/mailman/listinfo/insight-users
>  >  > >
>  >  > >
>  >  >
>  >  >
>  >  >
>  >  >  ----------------------------------------------------------------
>  >  >  Este mensaje ha sido enviado desde https://webmail.unav.es
>  >  >
>  >  >
>  >
>  >
>  >
>  >  --
>  >  Cheers,
>  >
>  >  Anja
>  >  _______________________________________________
>  >  Insight-users mailing list
>  >  Insight-users at itk.org
>  >  http://www.itk.org/mailman/listinfo/insight-users
>  >
>



-- 
Cheers,

Anja


More information about the Insight-users mailing list