Dear Stephan,
>Dear list,
>
>in a block design study with two stimulation tasks and one rest
>condition we find negative signal changes attributable to one of the
>tasks in a few cortex areas. In addition the whole CSF (not only parts of it)
>shows significant negative signal changes.
>Did anybody find something similar to this,
>and is there an explanation why the CSF should be less active under
>the task condition ?
>
No, I don't think there is an explanation why CSF should be "less active"
during tsak condition. I think what you see may be an artefact from the
global normalisation. By normalising the data with the global activity
(estimated as the average across all intracerebral voxels) then by
construction we constrain the estimated effects to sum to zero across all
voxels. This means that if there is an activation somewhere, especially if
that activation has a large magnitude and/or spatial extent, there will be
diffuse "deactivations" erroneously introduced elsewhere.
This problem stems from the average across all voxels not being a perfect
estimator of some underlying "global" change. Unfourtunately, simply
skipping the normalisation may also introduce bias. Hopefully there will be
better ways of performing the global normalisation available in the near
future.
Good luck Jesper
Jesper Andersson
Wellcome Dept. of Cognitive Neurology
12 Queen Square
London WC1N 3BG
phone: 44 171 833 7484
fax: 44 171 813 1420
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|