Dear Moran
your approach is overall the good one :-)
> Hi,
> I have new user questions for functional connectivity analysis in SPM
> I have time course (extracted from chosen VOIs), and I want to calculate correlation values between this area and each one of the brain voxels (I have rest fMRI data, i.e I'm only interesting in the functional connectivity and there is no model existent)
> After preprocessing I chose:
> Specific first level> Regressors= My VOI time course
> Next, after data estimation in the result section I choose: Type=t-contrast ; Contrast=1
you can also do -1 to see the negative correlation
> My questions are:
> 1. Is this right way to analyze my data?
> 2. This results with the following output files:
> beta_0001- if this in the image of estimated regression coefficients,
it is
> why the highest value (indeed located on my seed region) = 0.03 and not ~1?
> (I also got beta_0002, con_0001, mask, ResMS, RPV and spmT_0001 images.)
That's because your design matrix is actually bigger than what you see
... there is a filter built in and therefore the variance in your data
is correlated not only with the voi time course (although 0.03 seems
really low) but also with other regressors included in the filter -
therefore the beta values can't be one.
The beta_002 is for the second column in your design (the cst term). One
possible reason why the betas are so low could be the voi itself, you
may have taken a too large area and therefore the 1st eigen values over
voxels might be to different from the real data .. try a smaller region
to see if it improves the values
> If beta_0001 is not the most informative map, in which map should I look in order to see the calculated correlation value between my VOI and each one of the brain voxels and what it's scale stands for?
Betas are not the correlations coef although they reflect the strength
of the effect - to get r the standard way you need to compare the data
(SS total) to the model (SS effect) - but you can also convert the t
stat from spm_T images into correlation coef - the attached function
from Christian Gaser should do the trick
good luck
cyril
--
Dr Cyril Pernet,
fMRI Lead Researcher SINAPSE
SFC Brain Imaging Research Center
Division of Clinical Neurosciences
University of Edinburgh
Western General Hospital
Crewe Road
Edinburgh
EH4 2XU
Scotland, UK
[log in to unmask]
tel: +44(0)1315373661
http://www.sbirc.ed.ac.uk/cyril
http://www.sinapse.ac.uk/
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.
|