Hi, Steve.
Thank you for your response.
I'm beginner. Please forgive me even though my questions are vague or
definetly wronng.
1. About inverse transformation
I'd rather make clear to do inverse transformation of TBSS mask.
In FDT GUI, 'register' is performed after BEDPOST.
After 'register' process, I could get diff2standard.mat (standard space
registraiton to the 1mm MNI152)and also could have standard2diff.mat to
perform inverse transformation of TBSS mask. But, in 'register in FDT
GUI', the option of 'secondary image to apply transform' is not for the
inverse transformation of TBSS mask. Please let me know the method
(command?) which can transform from standard spaced- to individual
spaced-TBSS mask using standard2diff.mat.
Otherwise, because TBSS mask is already on 1mm MNI152 standard space, I
think, it could do using 'flirt linear registration in fsl main menu' like
below.
input image=avg152_T1_brain.hdr
reference image=subject's nodiff_brain.nii.gz
secondary image=TBSS mask image
cost function=correlation ratio? mutual information? (please let me
know which is appropriate)
2. about use of avwmaths++
>Also note that because the TBSS result is "on" the mean FA skeleton, you
may want to dilate it a little (using avwmaths++) before using it as a
mask.
--> Could you tell me more detail command to dilate individual space TBSS mask.
Thank you for much advance
Jeong
Stephen Smith wrote:
> Hi,
>
> At present we have not released a program for inverting the dof nonlinear reg warp files, that
> should hopefully be in the next release. The best way at present of doing what you want is to use
> the standard space registration in the FDT GUI to the 1mm MNI152 standard space image, and use
> this (inverted) to use your TBSS mask, in the normal way described in the FDT manual. This will
> therefore give affine registration from standard space back to native, but hopefully will be good
> enough. Also note that because the TBSS result is "on" the mean FA skeleton, you may want to
> dilate it a little (using avwmaths++) before using it as a mask.
>
> Cheers, Steve.
>
>
> On Sun, 21 Jan 2007 05:55:43 +0000, bs Jeong <[log in to unmask]> wrote:
>
>
>> I did TBSS and have gotten probability map of group difference (e.g.
>> tbss_maxc_tstat.nii.gz). To perform prob. tractography, I'd like to use this
>> probability map as seeding ROI. To do this, how can I transform inversly
>> this probability map from MNI space to individual space? This process may
>> use dof files which were calculated or mean_FA_skeleton.nii.gz which was
>> produced during tbss step 2 & 3.
>>
|