Dear Eli:
The volumes toolbox was written by Volkmar Glauche, but here is what I
believe is happening. You are probably using spm99. That version of spm
saved a compressed version of the raw data in a file called Y.mad. It only
saved data in which the F-test on the effects of interest exceeded the
significance as set by upper tail F-probability threshold. This is the
value fMRI_UFp in spm_defaults (default value of 0.001). The voxels for
which there is no saved data did not exceed the F-test at that location but
there can still be significant t-test results.
To avoid the problem of no saved data either set the spm_defaults value to
1 (fMRI_UFp = 1;) permanently or remember to modify the value using the
defaults button when you set up the statistics. Unfortunately you will have
to rerun the stats to get this data if it hasn't been saved. SPM2 avoids
this issue by always going back to the data volumes.
Regards,
Darren
At 03:51 PM 1/16/2005, Eli Packer wrote:
>Hello,
>Below is the messege I posted a few days a go. I am still struggling with
>the problem and can not solve it. I want to add that changing the treshold
>to various values does not help.
>
>Thanks,
>Eli
>
>
>*****************************
>Dir Sir/Madam,
>I am trying to extract FMRI data and having the following problem:
>I am using the toolboxes->volumes->multiple volumes->extarct maximum in
>order to find a voxel with best activity. Then, while using its
>coordinates
>to plot the data, the two following cases happen a lot:
>1) I get a window states that "No row data saved at this location: Data
>portions of plots unavailable at this location. Jump to the nearest
>voxel
>with saved raw data?"
>2) The coordinates are changed automatically (I guess that it happens
>when
>it finds a close voxels with data).
>I wonder why why this scenario occurs? if SPM finds a voxel with
>maximum
>activity, how comes he states that it has no data? Looking at the
>images
>implies that those maximum activated voxels have data, since they have
>color, otherthen white.
>Is there any way to fix this problem?
>Thanks a lot,
>Eli Packer
>Stony Brook University, USA
|