Thanks for your very quick response, John. When I ran a subset of my
data with files of different sizes, I received the following error
message. I will have to admit that I didn't start ripping through the
subroutines to find the error.
Running job #1
-----------------------------------------------------------------------
Running 'Run DARTEL (create Templates)'
Failed 'Run DARTEL (create Templates)'
Error using ==> dartel3
Incompatible dimensions.
In file "/usr/local/spm8/toolbox/DARTEL/spm_dartel_template.m" (v3172),
function "spm_dartel_template" at line 71.
The following modules did not run:
Failed: Run DARTEL (create Templates)
I was able to run through Dartel with the resliced data, but the
template was ... strange. There was a lot of complaining about NaN when
Dartel pushed through the data.
If you have any thoughts, I would be more than grateful.
pax,
Gustaf
0702033185
[log in to unmask]
----- Original Message -----
From: John Ashburner <[log in to unmask]>
Date: Friday, April 23, 2010 12:18 pm
Subject: [SPM] Reslicing data
To: Gustaf Mårtensson <[log in to unmask]>
Cc: [log in to unmask]
> The original image data can be any different size etc, as long as the
> spatially normalised data has the same dimensions. Also, if you
> plan to
> use Dartel, you would need to have "imported" data that is all in the
> same orientation, voxel size etc. The easiest way to achieve this
> is to
> use SPM to do the segmentation. In principle - provided all the
> headersare treated in a NIFTI compliant way - you could still use
> the new
> segmentation in SPM8 to obtain "imported" rc1 and rc2 files. These
> could then be Darteled together, and the resulting shape parameters
> usedto spatially normalise the results of the non-SPM segmented data.
>
> Best regards,
> -John
>
>
>
>
>
>
>
> Dear all,
>
> I have a set of data that we have segmented outside of SPM and would
> like to use
> to create a template and peform a parametric study. Unfortunately, the
> size of the
> study data matrices vary, while the voxel size is identical.
>
> I was wondering what the preferred method of reslicing (and aligning)
> the data would
> be (or is it padding that is needed)?
>
> Best regards,
>
> Gustaf Mårtensson
>
> Neuropediatric Unit, Karolinska Institutet
> Astrid Lindgren Children's Hospital, Q2:07
> SE-171 76 Stockholm, Sweden
>
> [log in to unmask]
> +46702033185
>
> --
> John Ashburner <[log in to unmask]>
>
>
|