#5 Spherical ROI in ROI Toolbox
Hi,
I may be missing something very simple here, but I
cannot seem to make the "define spherical ROI" option
on your toolbox work the way I want it to. My goal is to
take 10 ROIs from a previous paper, create a spherical
ROI for each, then use the toolbox to output timeseries
data for each in a new dataset (i.e., confirm the results
of a previous study using new data). My problem is that
while the "define spherical ROI" option works fine, it
does not spit out a con*.mat file. I'm also confused as
to why the "define functional ROI" option spits out a *.tal
file for each image, but the spherical ROI options does
not (i.e., what's the *.tal file contain?).
Am I just doing something simple wrong? Am I using
the program in a way it's not intended? Or is there a
bug here?
FYI -- I managed to spit out the 10 ROI maps just fine
individually. I even thought if I used image calc to put
them together into one image, copy it 20-30 times and
do a 1-sample t-test on it, then I might get an SPM.mat
file which I could then use in the "define functional ROI"
option to get the necessary con*mat file. No luck.
SPM99 bombs when trying to calculate the 1-sample t-
test on the resulting 16-bit ROI images. It adds the
option to implicitly mask the images (ignore zeros),
reads in the planes and planks, then crashes
saying "No inmask voxels - empty analysis!". This
makes me wonder how the spherical ROI mask that's
created is represented in the image.
Logged In: YES
user_id=95076
Hi - the spherical roi stuff should simply create a mask
image (with ones in the appropriate voxels in the image and
zeros everywhere else). the tal file is a vestige that is
not actually used by the roi programs, and will be gotten
rid of in the next revision. the con*.mat file is not
created for the spherical roi's, but I don't think that
should be a problem for any of the subsequent processes
other than viewing the roi.
The estimation of the data within the ROI should work using
either images derived with the function or spherical ROI
tools, and should not require a con*.mat file.
I hope this helps - if I've misunderstood your question
please let me know.
cheers
russ