Well, here's the story: We've analyzed the identical data using both SPM2
and SPM5. For SPM2, we used the code by Marko Wilke in which the SPM.mat
file is modified so that mask.img is replaced by some gray matter mask AFTER
the model is specified. That results in data that conform exactly with the
contour of our gray matter mask-- NOT just with the visual contour of the
BOLD signal map.
But, the identical data analyzed in SPM5 show only a subset of that exact
same specified gray matter (explicit) mask.
Perhaps Marco Wilke's code is doing something at a later stage? I suspect
that there is something else that needs to happen to accomplish our goal.
Thanks again for your interest in helping us. If you (or anyone else) has
other ideas, let us know!
David K
On Thu, 11 May 2006 08:45:49 +0200, Volkmar Glauche
<[log in to unmask]> wrote:
>On Wed, 10 May 2006, David Kareken wrote:
>
>> Thanks very much for responding. However, we just looked at that, and
>> this doesn't appear to be the problem. We submitted a mask that is mostly
>> 0's and 1.0s (some values ~0.6 around some edges). But, there are large
>> regions in which the mask clearly has 1.0s, and which are clearly excluded
>> in the analysis (i.e., in the mask.img). These also correspond to the usual
>> susceptibility-prone areas.
>>
>> Are there other things that you or anyone on the list can think of that
>> might account for this?
>
>Hmm... I am not sure it is a problem with the explicit mask then. Only
>voxels that survive ALL of the following criteria will be included in your
>analysis:
>- explicit mask greater zero
>- all values (over time/subject) are non-zero/non-NaN
>- all values (over time/subject) are greater than the mask threshold (if
> specified)
>- the statistics in this voxel survives the initial F-test.
>
>Most likely you will not have data for all timepoints/subjects in the
>excluded voxels.
>
>Volkmar
>
>--
>Volkmar Glauche
>-
>Department of Neurology [log in to unmask]
>Universitaetsklinikum Freiburg Phone 49(0)761-270-5331
>Breisacher Str. 64 Fax 49(0)761-270-5416
>79106 Freiburg
>=========================================================================
|