Hi Valeria,
> Hi all, we were wondering whether anyone can help us figure out this little
> problem we have.
>
> We are trying to run SPM on a simulated data set, which contains voxels
> with gaussian noise and several actived voxels with an ideal hemodynamic
> response. The dynamic range of data values is between 0 and 32767.
> Whenever we apply fMRI model on the data set in SPM99, it gives us the
> following error messages:
>
> -----------------------------------------------------------------------
> Warning: No inmask voxels - empty analysis!
> > In /home/wilkin/spm_spm.m at line 1076
> In /software/spm99/spm_fmri_spm_ui.m at line 619
> Design parameters : ...intrinsic autocorrelation
> Warning: Divide by zero.
> -----------------------------------------------------------------------
>
> We have tried several different setting and still get the same error. It
> appears that SPM99 applies some kind of scaling during the preprocessing
> steps. As the results, there is not voxel survive from the masking
> operation done by the script of "spm_spm.m".
>
> any suggestions of what can be causing this problem?
The following might explain the behaviour of SPM99 with respect to your
data:
CrLm(CrLm) = Y(i,CrLm) > xM.TH(i,1); %-Threshold (& NaN) mask
where
xM.TH is set to (g.*gSF) (in line 517 in spm_fmri_spm_ui.m)
As Darren pointed out, this is ok for real data, because (real)
intra-brain voxels tend to have all values over time greater than
xM.TH(i,1), where i is a specific time point. In your case what happens
(I guess) is that one or more voxel values in a time-series are below
xM.TH(i,1) such that the whole time-series is removed from analysis.
I can think of two ways around this:
The first (I personally would prefer this one) is to change line 517 in
spm_fmri_spm_ui.m to something like:
'TH', -inf,...
and the second is to remove the thresholding done in line 705 in
spm_spm.m
Hope this helps, Stefan
--
Stefan Kiebel
Functional Imaging Laboratory
Wellcome Dept. of Cognitive Neurology
12 Queen Square
WC1N 3BG London, UK
Tel.: +44-171-833-7478
FAX : -813-1420
email: [log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|