Posted by Florian Hauch on March 12, 2001 at 06:47:09:
Im doing some research on auditory stimulation in FMRI.
For analysation of signal response, I've made mask-datasets for
Brodman areas 41, 42 and 22 (from the talairach-daemon database).
To apply these masks on 3d+t datasets with 3Dmaskave I have to
convert the 3d+t to talairach space (I delete these files
immediatly after analyzation, because they have about 500MB each).
The original voxel-size from the EPI-sequence is about 2.2x2.2x8mm
so that every original voxel is warped to 40 voxels in talairach
space. When I view the corresponding graphs then all new voxels
generated from one original voxel have slightly different curves.
This seems logical for me, because all images of the timeseries
have been interpolated separately.
What is the consequence for the averaged signal response,
is the averaged timecourse of all talairach-voxels corresponding
to one original voxel equivalent to the original timecourse?
What effects can occur here?
Florian Hauch
Munich, Germany