The input images are all ok, as well as after registration. The error I get when I use run_first_all is that the registration is completely wrong. So, I used the flirt registration which worked fine, and then tried using first.

The funny thing is that I have three files. On one of them, it worked perfectly but on the other two it didn't. Now, I can't tell how they are different in any way. I tried copying the header of the one that worked into the other two and then running it again but no success. 

I uploaded the original files thru the support page. The reference number is: 266853 That is, the files without the registration. Again, any help is greatly appreciated! I've been stalling on this for a bit longer than I wanted. Thank you!

Pablo Martin

On Wed, Nov 28, 2012 at 6:13 PM, Mark Jenkinson <[log in to unmask]> wrote:
Hi,

This looks quite bad and unusual.
Are your input images OK?
Do they look fine when viewed in FSLView?
Is the registration OK?
Does the result of the registration look good when viewed on the standard brain using FSLView?

And what error message do you get when you use run_first_all?

If all of the above things seem OK, then upload the data and we'll have a look (instructions for the upload are on the Support page of the FSL website).

All the best,
        Mark



On 28 Nov 2012, at 15:03, Pablo Martin <[log in to unmask]> wrote:

> Hi,
>
> I'm having problems with run_first. I first tried run_first_all with not much success so decided on doing the registrations separately first. These worked out and now I'm trying to perform a segmentation. I'm trying to get the hippocampi.
>
> I encounter the following error:
>
> create shapeModel
> done creating shapeModel
> /usr/local/fsl/bin/run_first: line 164: 37686 Abort trap              ${FSLDIR}/bin/first -i $image -l $trmat -m $model -k $outname -n $modes $verbose $comExtras
>
> I have no idea what Abort trap could refer to and google so far hasn't helped. The algorithm doesn't output anything, not even some logs or whatever to see why it's crashing. This is how I run the command:
>
> run_first -i myfile.nii.gz -t myfile_to_std_sub.mat -n 50 -o output_name_L_Hipp -m /usr/local/fsl/data/first/models_336_bin/L_Hipp_bin.bmv
>
> This is how the documentation, if I understood it correctly, requires you to do it. Any ideas what is going on here?? How can I solve this?? Any help will be greatly appreciated!
>
> P.S. I'm running FSL 5.0.1 on a Mac 10.6. I'm pretty sure everything is installed properly and all the pathnames etc to the libraries are correct.
>