[Insight-users] Region Growing segmentation and viewing the results

Sergio Vera sergio.vera at alma3d.com
Thu Sep 15 09:29:53 EDT 2011


Hi John
Remember to reply also to the mailing list so your question may help others
in the future.

When I was referring to the value of the segmented object I meant the label
(pixel value) assigned to the region segmented by the region growing filter.
(1 or 255 usually)

Doing the segmentation in 3D or in 2D slices may affect the  segmentation
result, as the same object may be connected in a 3D image but it may appear
disconnected in a particular slice. In some cases performing slice by slice
segmentation might prevent unwanted leakage but on some other cases,
segmentation might be incomplete.
I'd say that 3D region growing is the first thing to try and only use slice
by slice on very specific cases.

Best regards


On Thu, Sep 15, 2011 at 3:05 PM, John Hiller <john.hiller53 at gmail.com>wrote:

> I've checked paraview, I'm able to see the first mha file (slice), it looks
> like it segmented that particular slice correctly. I'm reading my dicom
> files with the image type set to dimension 2 and writing each file into
> mha's with the same dimension, should I be reading and writing in 3D? I
> wouldn't think so since I'm processing each slice in turn so a 2D
> representation is necessary.
>
> I'm not sure what you mean by the value of my segmented object, what are
> you referring to as the segmented object (the whole mha series or one slice)
> and where can I find its value?
>
> I'm doing this slice by slice because I need to extract particular bones
> from the dicom representation of a skull to get a 3D visualization, but, I
> want the option to include or exclude fragments of these bones. The major
> influence for this choice is because of the interaction/HCI I've designed
> for the user.
>
> Thanks for the help
>
>
> On Thu, Sep 15, 2011 at 1:45 PM, Sergio Vera <sergio.vera at alma3d.com>wrote:
>
>> Hi
>>
>> First, check if the problem is related with segmentation or visualization.
>> Open your mha files with paraview/seg3d2 or whatever 3D viewer you are
>> used to. Check if the segmentation is ok. What value has your segmented
>> object? 1?, 255? create a isosurface to the appropiate value.
>> Once checked if you are segmenting the slices ok, move to the vtk side and
>> see if the fast marching is working properly or you are using the correct
>> isovalues.
>>
>> PS: any special reason to perform the segmentation slice by slice instead
>> of segmenting the whole volume in a single step?
>> Regards
>>
>> Sergio
>>
>> On Thu, Sep 15, 2011 at 2:36 PM, John Hiller <john.hiller53 at gmail.com>wrote:
>>
>>> Hi itk users
>>>
>>> I'm new to itk and vtk (sorry if this question doesn't apply to itk).
>>>
>>> Okay, so this is what I'm doing:
>>> 1) read a dicom series, I process each dicom slice individually
>>> 2) pick the segmentation points on a particular dicom slice
>>> 3) segment using confidence connected filter and save (or write to a file
>>> in other words) the result to a .mha file
>>> 4) Continuously do this for each dicom slice to get series of .mha files
>>> 5) Once all that is done, I attempt to use Marching cubes from vtk to
>>> render a 3D surface image from the .mha files.
>>>
>>> Everything works with no errors, its just that the vk window comes up
>>> black (showing nothing). I'm pretty sure the .mha files contain something
>>> (they're round about the same size as the original dicom file). I did the
>>> cout << reader->getOutput(), and it outputs information. I'm presuming that
>>> I am segmenting these slices incorrectly.
>>>
>>> Any advice would be helpful
>>> Thanks
>>> John
>>>
>>> _____________________________________
>>> 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
>>>
>>>
>>
>>
>> --
>> Sergio Vera
>>
>>  Alma IT Systems
>>  C/ Vilana, 4B, 4º 1ª
>>  08022 Barcelona
>>  T. (+34) 932 380 592
>>  www.alma3d.com
>>
>
>


-- 
Sergio Vera

 Alma IT Systems
 C/ Vilana, 4B, 4º 1ª
 08022 Barcelona
 T. (+34) 932 380 592
 www.alma3d.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20110915/def108ee/attachment.htm>


More information about the Insight-users mailing list