Dear Riccardo,
There are almost no visible anatomical features in your M0 image and this,
coupled with the low resolution and blurring, is making the registration *extremely*
difficult. I do not think that you will be able to get an intensity-based registration
method to work. What I recommend instead is to use the scanner coordinate
system for alignment instead, as this seems to work well. Just do:
flirt -in M0 -ref T1 -omat M0toT1.mat -out M0toT1 -usesqform -applyxfm
This seemed to give an acceptable registration, although not perfect.
However, I think this is the best you will be able to get with this data.
Also, please do not send data to our upload site in future without us requesting it
specifically. We don't have the space for everyone to load up datasets or the
time to look at them for every case.
All the best,
Mark
On 4 Jan 2012, at 17:55, Riccardo Navarra wrote:
> Dear FSL Experts,
> I need to register M0 scan to T1 (Six-digit reference number: 358158). I did many tests but I can't get a good registration. I notice a strange behavior with no clear translations and deformations.
> Please help! Thank you,
>
> Riccardo
>
|