help
help > RE: ROI-ROI fucntional connectivity matrix not symmetric
Apr 4, 2020 11:04 PM | Alfonso Nieto-Castanon - Boston University
RE: ROI-ROI fucntional connectivity matrix not symmetric
Hi Shabnam,
Sorry for the confusion, the Z matrices stored in those files are rectangular (as they are encoding the connectivity between your selected "seed" ROIs and all of the ROIs in your conn project), but the first and second dimensions of the Z variable are sorted in the same way, starting with your seed ROIs, so if you simply crop them to be square, e.g. using
Z = Z(:,1:size(Z,1),:);
the resulting matrices should be symmetric (up to machine-precision, and given your first-level analysis options; mainly bivariate-correlation measures)
Best
Alfonso
Originally posted by shabnamh:
Sorry for the confusion, the Z matrices stored in those files are rectangular (as they are encoding the connectivity between your selected "seed" ROIs and all of the ROIs in your conn project), but the first and second dimensions of the Z variable are sorted in the same way, starting with your seed ROIs, so if you simply crop them to be square, e.g. using
Z = Z(:,1:size(Z,1),:);
the resulting matrices should be symmetric (up to machine-precision, and given your first-level analysis options; mainly bivariate-correlation measures)
Best
Alfonso
Originally posted by shabnamh:
Hello,
I have been using the result of first level analysis (GLM, bivariate correlation, ROI-ROI, hrf weight) as the weighted connectivity matrix between ROIs for further functional connectivity network analysis. However, I noticed that these matrices (stored at e.g. resultsROI_condition1.mat as Z matrices for all subjects) are not symmetric. Aren't these supposed to show functional connectivity, as bivariate correlation, between the ROIs and be symmetric (Z= transpose(Z) for each subject)? Am I missing something here or looking at the wrong file?
Thanks much for your help,
Shabnam
I have been using the result of first level analysis (GLM, bivariate correlation, ROI-ROI, hrf weight) as the weighted connectivity matrix between ROIs for further functional connectivity network analysis. However, I noticed that these matrices (stored at e.g. resultsROI_condition1.mat as Z matrices for all subjects) are not symmetric. Aren't these supposed to show functional connectivity, as bivariate correlation, between the ROIs and be symmetric (Z= transpose(Z) for each subject)? Am I missing something here or looking at the wrong file?
Thanks much for your help,
Shabnam
Threaded View
| Title | Author | Date |
|---|---|---|
| shabnamh | Apr 2, 2020 | |
| Alfonso Nieto-Castanon | Apr 4, 2020 | |
| shabnamh | Apr 6, 2020 | |
| shabnamh | Apr 7, 2020 | |