Hello Volkmar,
That worked perfectly, thank you very much.
Just to elaborate on some of the steps for other readers:
in the batch, the first module is Named File Selector, with N sessions;
manually select the .nii file containing all the images for that session.
Next are the preprocessing modules. Say the first is Realign; its input
would be as usual: for each session, manually select all the files for
that session.
Then use dependencies all the way to the smooth step included.
Next use File Set Split. Its input is the dependency on all smoothed images.
Its output, for each session, is a row vector that gives the positions
of the images
in the list of all smoothed images. For example, with 120 scans per session,
the way to specify the output for first session is [1:120], for the
second session: [121:240], etc.
Next are the stats steps: the model specification step takes as input
the dependency
from the File Set Split step. The other stat steps are as usual.
Again, thanks a lot,
Philippe
On 12 April 2010 03:23, Volkmar Glauche
<[log in to unmask]> wrote:
> Dear Philippe,
>
> the smooth processing step will not distinguish between sessions, and
> therefore it returns a single list of file names. To split them up into
> files per session again, you may want to do the following:
> At the top of your batch, insert a "Named File Selector" from BasicIO. Add
> as many file sets as there are sessions. Pass the file sets on to realign,
> slicetiming etc. (Note, that the "Named File Selector" is not constrained to
> NIfTI images, you may need to filter images manually in the selection
> dialog.)
> After the smooth step, insert a "File Set Split". Enter the smoothed images
> as files, and create as many output filesets as there are sessions. For each
> of them, enter the selection index from the corresponding "Named File
> Selector" file set.
>
> Hope this helps,
>
> Volkmar
>
> <https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=SPM;661f7458.1004>
>
|