[Insight-users] NoiseImageFilter for 3D images

[Mirco Serra]

Hi guys,
thanks for your help. It works fine now, I had to use short instead of unsigned short as Bradley suggested!
Many many thanks,
Mirco

Date: Fri, 16 Aug 2013 09:36:25 -0400
Subject: Re: [Insight-users] NoiseImageFilter for 3D images
From: bill.lorensen at gmail.com
To: blowekamp at mail.nih.gov
CC: mirco9091 at hotmail.it; insight-users at itk.org

CT data is usually signed. Try short instead of unsigned short.

On Fri, Aug 16, 2013 at 9:19 AM, Bradley Lowekamp <blowekamp at mail.nih.gov> wrote:

Hello Micro,
I don't regularly work with CT, so I don't know for certain. What is the pixel type in your file?

http://www.itk.org/Wiki/ITK/Examples/IO/ReadUnknownImageType
Also why do you say you image is all zeros? Did you inspect it in am image viewer? Perhaps all the values were just too low to be seen?

It may be best to have the output of the filter be at least of float, since its computing the standard deviation will be low in homogeneous regions which may be causing you to thing it's zero.

You are correct that in 3d a radius of 2 aka [2,2,2] will yield a neighborhood with a size of [5,5,5].  This has to do with how the Neighborhood operators and iterator work. You can find some example which may help you understand the behavior better here:

http://www.itk.org/Doxygen/html/examples.htmlhttp://www.itk.org/Wiki/ITK/Examples

Brad

On Aug 16, 2013, at 9:02 AM, Mirco Serra <mirco9091 at hotmail.it> wrote:


Hello,
Thanks for your answer!
I am using unsigned short as image type both for the input and the output image. Is this the correct type for 3D CT images? Also, would you mind explaining me the way the radius for the local region is selected for 3D images? In 2D I know that a radius value of 1 corresponds to a 3x3 region, a radius of 2 is a 5x5 region and so on. Is it the same in 3D, i.e. a value of 1 is a region of 3x3x3?

Many thanks,
Mirco

Subject: Re: [Insight-users] NoiseImageFilter for 3D images
From: blowekamp at mail.nih.gov

Date: Fri, 16 Aug 2013 08:48:56 -0400
CC: insight-users at itk.org
To: mirco9091 at hotmail.it


Hello,
What are your template parameters to the filter i.e. the image types used?
I just tried this filter in SimpleITK with a couple different images and it seems to work fine.

Thanks,Brad
On Aug 16, 2013, at 5:05 AM, Mirco Serra <mirco9091 at hotmail.it> wrote:


Hi guys,
I am trying to use the NoiseImageFilter class in order to get the local standard deviation of a 3D DICOM/nrrd dataset. The filter works nicely with 2D images, but when I try to use for 3D datasets it does not work properly and it returns an image with only zeros as intensity values. Do you have any idea why this could happen? I struggled with it a lot but I could not find a solution! Is the filter for 3D images too?

All the best and thank you very much,
Mirco _____________________________________
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





_____________________________________

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




-- 
Unpaid intern in BillsBasement at noware dot com

 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20130817/b8f60e60/attachment.htm>