[Insight-users] Brain Segmentation of inhomogeneous images

[Karthik Krishnan]

As Luis mentioned you should take a look at the bias field correction
application in InsightApplications.

Given that you are using a bayesian classifier, you already have a rough
estimate of the tissue statistics (priors). The bias field correction
application uses precisely these (in this specific case, it assumes a
gaussian model, using the mean and standard deviation of each tissue class)
to estimate the correcting polynomial. So the application should be well
suited to your problem.

On Wed, Feb 11, 2009 at 5:05 AM, <a.bongers at mediri.com> wrote:

> Hi,
>
> I am trying to use a Bayes Classifier to segment white ang gray brain
> matter from MRI images. Works O.K. so far, but I have the problem that many
> 3D images have varying gray-values e.g. in within the white matter. Thats
> probably due to RF inhomognieties from the coil or the like.
> The problem is that due to these slowly varying intensities the
> segmentation mis-calssifies GM and WM e.g. in the back of the brain.
>
> Has anybody a helpful idea to get better results ?
> I was thinking about hig-pass filtering the images to homogenize the
> intensities.
> Is there a high- or bandpass filter in itk ?
>
> Thanks a lot!
>
> André
> _____________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.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
>



-- 
Karthik Krishnan
R&D Engineer,
Kitware Inc.
Ph: 518 371 3971 x119
Fax: 518 371 3971
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20090215/58065790/attachment.htm>