Hi Andreas,
thanks for your help.
I don't think my explanation was very clear.... but I think that thanks to your
help and Sascha, I am arriving at a solution.
I have quite a reasonably complex pipeline of ITK filters, that all need to
work on a 3D image. The first filter is basically a thresholding. But rather
than threshold the 3D volume, which can be a bit slow, I wanted to just
threshold the 2D slices that were visible in the viewer. However, this is
probably not worth the effort of extracting individual 2D slices, putting them
in the DataStorage etc. So, I will threshold the whole 3D volume, and then any
view can take a slice through the resulting volume and display it as
appropriate.
I will use basically this suggestion:
https://gist.github.com/1113999
I still need a control (a slider currently) to select an axial slice, which
will simply remove voxels from the segmented image that are below a certain
slice. i.e. remove the neck region.
So, I still need to know how to identify which orientation is Axial, as
mentioned here:
http://sourceforge.net/mailarchive/forum.php?thread_name=99573E43-4215-4E4E-A31F-485F50CF6F16%40live.ucl.ac.uk&forum_name=mitk-users
Many thanks
Matt
------------------------------------------------------
Matt Clarkson Ph.D.
CMIC Software Manager
Senior Research Associate
[email protected]<mailto:[email protected]>
Centre For Medical Image Computing
http://cmic.cs.ucl.ac.uk/staff/matt_clarkson/<http://cmic.cs.ucl.ac.uk/>
Tel: 020 7679 0221
Fax: 020 7679 0255
Dementia Research Centre
http://dementia.ion.ucl.ac.uk/
Tel: 08451 555 000 ext. 723653
Fax: 020 7676 2066
------------------------------------------------------
On 22 Jul 2011, at 07:47, Fetzer, Andreas wrote:
Hi Matt,
1. how do I access those slices?
To access / extract normally oriented slices MITK provides you the
mitkExtractImageFilter. By setting the slice dimension you can tell the filter
to extract transversal/coronal/sagittal slices. If you need an example of how
to use this filter you can look into the mitkSegTool2D at line 180 -190. If you
want to write the processed slice back into your image volume you should use
the mitkOverwriteImageFilter, which is used similar to the ExtractImageFilter.
Regarding the outlining: What exactly do you want to do? E.g. you could extract
single contours out of slices or you could create a complete 3D surface out of
your segmentation which is displayed as a 2D contour in the 2D windows and
hence outlines your segmentation. Another option is to tell MITK that is should
outline your segmentation. In this case you do not have filled contours but
just the edges of them displayed in the 2D windows.
2.
I am not sure if I understood exactly what you want to do, so please correct me
if I got in the wrong direction. Basically you will have 2 objects in the
datamanager: your original volume and the thresholded version, i.e. your
segmentation. If you want to cut off certain areas of your segmentation of
course you could create a new image and put it into the datamanager either. But
you can also just delete the appropriate pixel values in your segmentation. The
last version should be more efficient, if you don`t need the deleted image
information in a later step.
Why do you want to update the geometry manually? Normally you don`t have to do
that. How exactly do you edit your image?
I hope this helps you.
Best regards
Andreas
-----Ursprüngliche Nachricht-----
Von: Clarkson, Matt [mailto:[email protected]]
Gesendet: Dienstag, 19. Juli 2011 16:35
An: [email protected]<mailto:[email protected]>
Betreff: [mitk-users] Question on thresholding and outlining a 2D slice
Hi there,
please can I have thoughts on how best to do the following task:
Imagine I have loaded a volume, and I'm going to step through a sequence of
buttons/sliders that threshold an image, then apply erosions, dilations, a few
other ITK filters, and we need to allow manual editing.
1. The first thing is to select thresholds, and an axial slice, below which the
image is always thresholded. To keep things snappy, I don't want to be applying
thresholds to a 3D image. So, if my orthogonal view planes show 3 slices
(axial, coronal, sagittal), how do I access those slices, so I can threshold
just them slices, and overlay the outline of the thresholded volume ontop of
those same slices?
2. I also need to apply an "Axial cutoff", so I need to select an axial slice,
below which the image is thresholded to zero regardless of the minimum and
maximum limits set on the slider. THis is so that you can remove the neck
easily on an MR image. For this I would need to compute if a voxel had a voxel
index below/above the chosen slice, so if in step 1, I have access to the 2D
slice I only need a 2D filter, but where would I put this stuff? Do I create 3
new data objects in the DataManager, and update their geometry manually,
keeping them in sync with the current slice selection, and let the usual
pipelines render them? Or is there a better way?
(i hope this makes sense).
Once the threshold and axial cutoff slice is chosen, the rest of the
application runs on the 3D image, so this is fairly straight-forward in MITK
terms, as we simply do image processing in ITK.
Thanks
Matt
------------------------------------------------------------------------------
Magic Quadrant for Content-Aware Data Loss Prevention Research study explores
the data loss prevention market. Includes in-depth analysis on the changes
within the DLP market, and the criteria used to evaluate the strengths and
weaknesses of these DLP solutions.
http://www.accelacomm.com/jaw/sfnl/114/51385063/
_______________________________________________
mitk-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mitk-users
------------------------------------------------------------------------------
BlackBerry® DevCon Americas, Oct. 18-20, San Francisco, CA
The must-attend event for mobile developers. Connect with experts.
Get tools for creating Super Apps. See the latest technologies.
Sessions, hands-on labs, demos & much more. Register early & save!
http://p.sf.net/sfu/rim-blackberry-1
_______________________________________________
mitk-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mitk-users
