Tell me more ×
Cross Validated is a question and answer site for statisticians, data analysts, data miners and data visualization experts. It's 100% free, no registration required.
up vote 2 down vote favorite
1

I want to analyze functional MRI (fMRI) data as follows:

  1. I am comparing the brain networks of two groups of subjects (patients and matched).

  2. For each subject I have a correlation (Pearson) matrix with the correlations between multiple brain regions (above 50000 correlations).

  3. I want to get a single measure to represent every subject (maybe the mean of correlations of each subject network). What is the most suitable measure? Should I use Fisher transformation to average the correlation matrix?
  4. I assume that the variance of the two groups is different and that the variance in each group is similar: How can check that without doing multiple comparisons?
share|improve this question
1  
Do you really want to do an FMRI analysis with absolutely no chance of localization of the functional differences? – John Dec 27 '11 at 15:02
I am comparing voxels correlations within specific regions of interest, so it is localized. – gid3a1 Dec 27 '11 at 15:17
ah, so you just want to see which group's region X better correlates. – John Dec 27 '11 at 15:34
Several regions, but yes. – gid3a1 Dec 27 '11 at 15:38
Whether it's physically the same region is irrelevant, only whether you're treating the same functionally... since that's what you're studying. – John Dec 27 '11 at 16:06
show 1 more comment

1 Answer

up vote 1 down vote accepted

Yes, z-transform to average the functional regions within S and then compare them as you'd normally compare correlations between groups... is your simplest path. Otherwise, I thought FSL had a facility to handle such comparisons of the magnitude of an effect with multi-level modelling.

share|improve this answer
I'm using brain voyager. Anyway I wanted to know how to do it theoretically. Thank you. – gid3a1 Dec 27 '11 at 16:46

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.