Automate segmentation of ventricles from Brain MRI

classic Classic list List threaded Threaded
3 messages Options
919
Reply | Threaded
Open this post in threaded view
|

Automate segmentation of ventricles from Brain MRI

919
This post has NOT been accepted by the mailing list yet.
I have about 50 MRI series (T1, T2, Flair and T1 volumetric) that I would like to segment out the ventricles to generate 3D models and to calculate the volumes.  I would appreciate any input as to how best to go about doing this in a manner that is as autonomous as possible.

One method I was considering was using one of the brain atlases to identify and label all of the anatomy and then selectively export just the ventricles for each MRI. At this point I am having some difficulty with using the BRAINS fit to non rigidly register the atlas a subjects MRI in such a way that the CSF space or ventricles are correctly labeled.

Any help would be great with this process or if someone has another suggestion such as trying to automate the segmentation of just the ventricles.  I did see in one of the tutorials where you can use a module to label the CSF space, skull, white mattter and gray matter, maybe this would be another step
Reply | Threaded
Open this post in threaded view
|

Re: Automate segmentation of ventricles from Brain MRI

Harary, Maya
I would use the T2 image (no atlas needed), Use the Editor module > Use the paintbrush to mark the ventricles (one brush stroke is enough, it is just to give the program a reference) > Change the color/label of the brush to something else (default is Green/”tissue”) > Mark the periventricular area as reference > Click the GrowCutEffect tool (2nd icon row, 3rd from left), click Apply

This should segment the ventricle separate from the periventricular tissue in all slices of the series you have loaded. You should scroll through to ensure that is the case, and tweak by hand the errors

To quantify the volume, go to the Quantification module > Label statistics > Select your volumes, click apply > this should give you the volumes

Let me know if you have any questions/if it works out ok.

Maya

On 3/2/17, 19:47, "slicer-users on behalf of 919" <[hidden email] on behalf of [hidden email]> wrote:

    I have about 50 MRI series (T1, T2, Flair and T1 volumetric) that I would
    like to segment out the ventricles to generate 3D models and to calculate
    the volumes.  I would appreciate any input as to how best to go about doing
    this in a manner that is as autonomous as possible.
   
    One method I was considering was using one of the brain atlases to identify
    and label all of the anatomy and then selectively export just the ventricles
    for each MRI. At this point I am having some difficulty with using the
    BRAINS fit to non rigidly register the atlas a subjects MRI in such a way
    that the CSF space or ventricles are correctly labeled.
   
    Any help would be great with this process or if someone has another
    suggestion such as trying to automate the segmentation of just the
    ventricles.  I did see in one of the tutorials where you can use a module to
    label the CSF space, skull, white mattter and gray matter, maybe this would
    be another step
   
   
   
    --
    View this message in context: http://slicer-users.65878.n3.nabble.com/Automate-segmentation-of-ventricles-from-Brain-MRI-tp4031932.html
    Sent from the slicer-users mailing list archive at Nabble.com.
    _______________________________________________
    slicer-users mailing list
    [hidden email]
    http://massmail.spl.harvard.edu/mailman/listinfo/slicer-users
    To unsubscribe: send email to [hidden email] with unsubscribe as the subject
    http://www.slicer.org/slicerWiki/index.php/Documentation/4.3/FAQ
   

_______________________________________________
slicer-users mailing list
[hidden email]
http://massmail.spl.harvard.edu/mailman/listinfo/slicer-users
To unsubscribe: send email to [hidden email] with unsubscribe as the subject
http://www.slicer.org/slicerWiki/index.php/Documentation/4.3/FAQ
Reply | Threaded
Open this post in threaded view
|

Re: Automate segmentation of ventricles from Brain MRI

Andras Lasso-2

Yes, this approach should work well. Also, seeds can be generated automatically by registering an atlas image (see details in our paper: http://perk.cs.queensu.ca/contents/quantification-intraventricular-blood-clot-mr-guided-focused-ultrasound-surgery).

 

Also, I would strongly recommend using the “Grow from seeds” effect in the new Segment editor module. It is about 10x faster than growcut effect in the old editor and allows quick preview/update of results, which is about 100x faster.

 

Andras

 

From: [hidden email]
Sent: Thursday, March 2, 2017 8:09 PM
To: [hidden email]; [hidden email]
Subject: Re: [slicer-users] Automate segmentation of ventricles from Brain MRI

 

I would use the T2 image (no atlas needed), Use the Editor module > Use the paintbrush to mark the ventricles (one brush stroke is enough, it is just to give the program a reference) > Change the color/label of the brush to something else (default is Green/”tissue”) > Mark the periventricular area as reference > Click the GrowCutEffect tool (2nd icon row, 3rd from left), click Apply

This should segment the ventricle separate from the periventricular tissue in all slices of the series you have loaded. You should scroll through to ensure that is the case, and tweak by hand the errors

To quantify the volume, go to the Quantification module > Label statistics > Select your volumes, click apply > this should give you the volumes

Let me know if you have any questions/if it works out ok.

Maya

On 3/2/17, 19:47, "slicer-users on behalf of 919" <[hidden email] on behalf of [hidden email]> wrote:

    I have about 50 MRI series (T1, T2, Flair and T1 volumetric) that I would
    like to segment out the ventricles to generate 3D models and to calculate
    the volumes.  I would appreciate any input as to how best to go about doing
    this in a manner that is as autonomous as possible.
   
    One method I was considering was using one of the brain atlases to identify
    and label all of the anatomy and then selectively export just the ventricles
    for each MRI. At this point I am having some difficulty with using the
    BRAINS fit to non rigidly register the atlas a subjects MRI in such a way
    that the CSF space or ventricles are correctly labeled.
   
    Any help would be great with this process or if someone has another
    suggestion such as trying to automate the segmentation of just the
    ventricles.  I did see in one of the tutorials where you can use a module to
    label the CSF space, skull, white mattter and gray matter, maybe this would
    be another step
   
   
   
    --
    View this message in context: http://slicer-users.65878.n3.nabble.com/Automate-segmentation-of-ventricles-from-Brain-MRI-tp4031932.html
    Sent from the slicer-users mailing list archive at Nabble.com.
    _______________________________________________
    slicer-users mailing list
    [hidden email]
    http://massmail.spl.harvard.edu/mailman/listinfo/slicer-users
    To unsubscribe: send email to [hidden email] with unsubscribe as the subject
    http://www.slicer.org/slicerWiki/index.php/Documentation/4.3/FAQ
   

_______________________________________________
slicer-users mailing list
[hidden email]
http://massmail.spl.harvard.edu/mailman/listinfo/slicer-users
To unsubscribe: send email to [hidden email] with unsubscribe as the subject
http://www.slicer.org/slicerWiki/index.php/Documentation/4.3/FAQ

_______________________________________________
slicer-users mailing list
[hidden email]
http://massmail.spl.harvard.edu/mailman/listinfo/slicer-users
To unsubscribe: send email to [hidden email] with unsubscribe as the subject
http://www.slicer.org/slicerWiki/index.php/Documentation/4.3/FAQ