Hi Jesper,
I meant that the brain extracted images look good.
I tried your fslmerge command, but got this error "Error in size-match along
non-concatenated dimension".
This may be because I cropped the original images with fslroi to remove the
neck, followed by brain masking, and running them through FAST to get the
"restore" image so as to remove the bias field inherent in our 3T images.
The restore images were then fed into fslvbm_1_bet.
I did ask the forum if the different sized images would be a problem for
fslvbm, and was told that it wouldn't be.
There are a lot of images for each subject with "struc_brain" in their
titles. I presume you would like to see the unsegmented ones.
Since I cannot fslmerge then, would the template_4D_GM file be useful to
you?
Thanks
----- Original Message -----
From: "Jesper Andersson" <[log in to unmask]>
To: <[log in to unmask]>
Sent: Monday, June 07, 2010 8:04 PM
Subject: Re: [FSL] FSL-VBM vs Freesurfer
> Hi again,
>
>> I think the images I have fed into the processing pipeline look very
>> good. Is it possible to send you some examples for your comment? Which
>> ones would be most helpful for review?
>
> what do you mean by "fed into"? Do you mean your original images, or do
> you mean the *_brain.* images you get after the fslvbm_1_bet step? If you
> have not checked the *_brain.* images I suggest you do so, looking for
> signs of brain being eroded or of non-brain being left in.
>
> If that looks good your next step should be to have a look at the
> template_GM_4D file to see how the registration has succeeded.
>
> If you still cannot find the source of the problem you can put all the
> *_brain.* images into a 4D file with the command
>
> fslmerge -t ForJesper `echo *_brain.*`
>
> If the resulting ForJesper file is not too large you can then upload it
> to
>
> http://www.fmrib.ox.ac.uk/cgi-bin/upload.cgi
>
> and email me the reference number. If it is very big you can change the
> fslmerge command above to only include a few of the *_brain.* files.
>
> Jesper
|