Dear experts,
I'm cutting in on this thread because I have a similar problem. I want to register a thresh_zstat image (an output component from MELODIC) to a DMN template already in MNI space. I was using FLIRT but after reading this post I'm assuming I should be using applywarp instead? Even after reading the practical on handling warps I'm still not quite sure what the --warp input should be. Should I be applying FLIRT or FNIRT before applywarp?
Thank you in advance!
Helen
________________________________________
From: FSL - FMRIB's Software Library [[log in to unmask]] On Behalf Of Mark Jenkinson [[log in to unmask]]
Sent: Friday, January 18, 2013 2:43 AM
To: [log in to unmask]
Subject: Re: [FSL] Normalising ROI to template
Hi,
Just use applywarp for the ROIs - you should never run FLIRT or FNIRT to do registration based on ROIs (it is OK to use flirt in the applyxfm mode, but not to perform a registration).
You need to make sure you have the right syntax (i.e. --ref=MNI, not -ref=MNI) and you probably want to set the output datatype to float (--datatype=float) and rethreshold the output with fslmaths, if you want the output to be a binary mask.
All the best,
Mark
On 17 Jan 2013, at 12:39, Verena Rozanski <[log in to unmask]> wrote:
> Dear experts,
>
> my aim is to normalise ROIs in anatomical space to the MNI template. I have already normalised the individual T1 to MNI space performing FLIRT, FNIRT and WARP.
> Is it possible to use the command:
>
> applywarp –ref= MNI –in=ROI_inT1 –warp=T1fnirt2MNI -out= ROI_in MNI ?
>
> Or do I have to perform the FLIRT and FNIRT steps for each ROI using applyxfm?
>
> Thank you for your support!
>
> Kind regards,
>
> Verena
