Thanks for your answer Steve. I can open up the Niftis with FSLview
and see every volume. Maybe the conversion via dcm2nii brought some
errors. How do I check if it is really corrupt? I will try it with
another converter.
Matthias
On 17 Dec 2013, at 10:20:38, Stephen Smith wrote:
> Hi -= you can see from your log (the first part of “Initialisation”)
> that even your 4D fmri input data > is problematic - either the file
> is corrupt or unreadable I guess?
>
> Steve.
>
> On 17 Dec 2013, at 01:28, Matthias Heil wrote:
>
> > [...]
> > I have enough disk space, swap (22GB) and ram (4GB). I also tried
> the > analysis with a smaller nifti (only the first 100 dicoms) ->
> same errors.
> >
> > [...]
> > I don't know if its important: I'm working on multiband images (factor 6 in
> > this case). Melodic did work fine before on other data. It's also
> the > first case
> > in which I use the BBR, no resampling and Nonlinear
> Registration/fnirt, and 2
> > resting state sessions of the same person (so same structural/T1
> image; I did
> > choose multi-session temporal concatenation). Brain-extracted and
> > non-brain-extracted do exist with the right name. At the moment I'm only
> > working with the GUI.
> >
> > Thanks from Dresden at night with no snow,
> > Matthias
|