In my experience,
Skull-stripping using BET or manual procedures is done on the volumes prior to quantification (segmentation, lesion localization, perfusion-DTI, relaxation,..) this reduces the number of "unknowns" and provides more faster and accurate registrations.
1. I see no reason why BET should not work on spatially normalized volumes in analyze or any other format. I am assuming the contrast between WM-GM-CSF are still ok as in the standard MNI-T1w provided in MRIcro.
2. I would use MRIcro and if you encounter problems with the BET called in MRIcro GUI. May utilize ROI-VOI smoothing options (CTRL-q, multiple times) and then create an intensity binary ROI-mask, save and then modulate by this file consquently and edit as needed as this is a Universal mask that is only needed once.
3. On the other hand, the thresholded Unified segmentation results of the template can be used so that a binary mask may be returned (This is what Bas just suggested).
Khader M Hasan, PhD
Associate Professor of Radiology
MSE 168, Tel 713 500 7690 (FAX 713 500 7684)
University of Texas Health Science Center at Houston
Medical School
Diagnostic and Interventional Imaging
Magnetic Resonance Imaging Research Division
Diffusion Tensor Imaging Lab, Tel 713 500 7683
http://www.uth.tmc.edu/radiology/faculty/khader-m-hasan/index.html
________________________________________
From: SPM (Statistical Parametric Mapping) [[log in to unmask]] On Behalf Of Chaleece Sandberg [[log in to unmask]]
Sent: Thursday, May 03, 2012 11:47 AM
To: [log in to unmask]
Subject: [SPM] skull-stripping AFTER normalization
Hi all,
I have lesioned brains and I use lesion masking to perform normalization. I would like to be able to use rendered images of the skull-stripped brains for presentation, but when I run the brain extraction tool in MRIcroN, it only works with the original nifti structural files, not the normalized images. Does anyone know why this is so? Does normalization result in an image that's no longer able to be segmented? Is there any way to skull-strip after normalization?
Thanks,
Chaleece
|