Dear Elizabeth,
At 10:45 22/11/1999 -0800, Dr Elizabeth Sowell wrote:
| I am running a PET type analysis compare groups 1 scan per
| subject (actually structural grey matter maps). I enter the scans
| for each group (n=10 each group), then selecting
|
| no global normalization
| no grand mean scaling
| threshold masking: absolute 0
| global calculation: omit (but I'm still not sure what this means)
As you are not doing any global normalisation or grand mean scaling,
the global mean is not needed, and you are given the option to omit its
calculation. Nothing sophisticated!
----------------
| After that, I get the error message
|
| SPM99b: spm_spm_ui (v2.25) 10:27:06 - 22/11/1999
| ========================================================================
| Mapping files : ...done
| ??? Error while evaluating uicontrol Callback.
|
| Inside the SPM windows I see the "clock" as though it is trying to process
| something
|
| When I hit return in the matlab window immediately after the error, I get the
| following message
|
| >> ??? Error using ==> spm_spm_ui
| images do not all have same orientation & voxel size
|
| My images are all, as I said, structural binary grey matter images which I
| created outside SPM, then imported and smoothed in SPM. The headers all show
| the same dimensions etc. and I can display all the images in SPM and they
| look fine. I edited the headers from SPM to make the origin the AC for each
| subject.
|
| Sample header:
|
| bits/pixel=8
| x_dim=181
| y_dim=217
| z_dim=181
| x_size=1.000000e+00
| y_size=1.000000e+00
| z_size=1.000000e+00
|
|
| These are the details on the typical analysis that I ran frequently with
| SPM96, and would desperately like to run with SPM99.
|
| Any more thoughts on this will be appreciated.
Note that if there is a *.mat alongside the image, then the orientation
and voxel sizes are taken from this *.mat file regardless of the values
set in the Analyze file header.
You can look at the orientations and voxel sizes SPM is using with the
following code (copy and paste into the MatLab command window):
%-----------------------------------------------------------------------
P = spm_get(Inf,'*.img','Select image files')
VY = spm_vol(P);
ImDims = cat(1,VY.dim) %-Image X, Y & Z dimensions (in voxels) & datatype
%-All images must have the same dimensions
Orient = cat(3,VY.mat) %-Orientation matrices (stacked in 3D array)
%-(See spm_matrix)
%-All images must have the same orientation matrix
%-To find the images where the orientations differ:
tmp = find(any(diff(reshape(cat(3,VY.mat),16,length(VY)),1,2),1))
% tmp contains the indices of images with different orientations to the next one
for i = tmp
fprintf(['Images %d & %d differ in orientation/voxel size\n',...
'\t%d:\t%s\n\t%d:\t%s\n'],i,i+1,i,VY(i).fname,i+1,VY(i+1).fname)
VY(i).mat
VY(i+1).mat
end
%-----------------------------------------------------------------------
Hope this helps,
-andrew
+ - Dr Andrew Holmes mailto:[log in to unmask]
| Robertson Centre for Biostatistics ( ,) / _)( ,)
| Boyd Orr Building, University Ave., ) \( (_ ) ,\
| Glasgow. G12 8QQ Scotland, UK. (_)\_)\__)(___/
+ - http://www.rcb.gla.ac.uk/
----------------
| > At 15:38 05/10/99 -0700, Dr Elizabeth Sowell wrote:
| > | I am for the first time running an analysis with SPM99 and having
| > | some problems. I would revert to SPM96 just to get quick results
| > | from the analysis, but my Matlab 4.2 lisence expired and I can't get
| > | SPM96 to run ever again!! Following are my 2 questions.
| > |
| > | 1. I am running a PET type analysis compare groups 1 scan per
| > | subject (actually structural grey matter maps). I enter the scans
| > | for each group, and then go through the design options up to global
| > | calculations where I have the option of omitting, user specified or
| > | mean voxel value. What is this option for and how does it relate to
| > | SPM96 design options?
| > |
| > | 2. When I select omit in the global calculations option (because
| > | I'm not sure what it is), I get the following error,
| > |
| > |
| > | SPM99b: spm_spm_ui (v2.25) 15:30:20 - 05/10/1999
| > | ======================================================================
| > | Mapping files : ...done
| > | ??? Error while evaluating uicontrol Callback.
| > |
| > | and I am not given the option to enter contrasts etc. I got the
| > | same error when I selected the mean voxel value option in this
| > | window as well. How can I get through this analysis?
| > |
| > | I tried looking in the mail group archives, but was not able to
| > | determine the problem for each of the questions above.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|