help
help > RE: ROI-to-ROI individual connectivity
Aug 22, 2013 06:08 PM | Julia Berezutskaya
RE: ROI-to-ROI individual connectivity
Dear Dan Zhu,
I did exactly what is said in FAQ:
For each of my experimental conditions I have a .mat file in the first level analysis folder named resultsROI_ConditionX (there are also separate .mat files for each subject, but I think it is less comfortable). When I load it to MATLAB the resulting structure has several fields one of which is called Z and it is a seed x ROI x subj matrix. The seeds correspond to those stored in the field names, the ROIs do to those stored in the field names2. The values Z(seed, ROI, subj) are the Fisher-transformed correlation coefficients, i.e. the values for connectivity between the seed and the ROI.
E.G.
I have 3 seeds (mPFC, ACC and PCC), the same ROIs and 27 subjects. My Z matrix then is 3 x 3 x 27. To extract the correlation values between mPFC and ACC for ConditionX I simply do that:
Z matrix is symmetric about the diagonal of NaN.
I am not sure if this is clearer than FAQ, but I hope it will be helpful.
I did exactly what is said in FAQ:
This file will contain a matrix Z with the
ROI-to-ROI connectivity values (Fisher-transformed correlation
coefficients). In particular the value Z(i,j) will contain the
connectivity between source ROI 'i' and target ROI 'j'.
For each of my experimental conditions I have a .mat file in the first level analysis folder named resultsROI_ConditionX (there are also separate .mat files for each subject, but I think it is less comfortable). When I load it to MATLAB the resulting structure has several fields one of which is called Z and it is a seed x ROI x subj matrix. The seeds correspond to those stored in the field names, the ROIs do to those stored in the field names2. The values Z(seed, ROI, subj) are the Fisher-transformed correlation coefficients, i.e. the values for connectivity between the seed and the ROI.
E.G.
I have 3 seeds (mPFC, ACC and PCC), the same ROIs and 27 subjects. My Z matrix then is 3 x 3 x 27. To extract the correlation values between mPFC and ACC for ConditionX I simply do that:
connData = load('connName.mat'); %load the
.mat file
corrVals(:, x) = squeeze(connData.Z(1,2,:)); %x-condition number, 1-mPFC, 2-ACC, :-for all the subjs
corrVals(:, x) = squeeze(connData.Z(1,2,:)); %x-condition number, 1-mPFC, 2-ACC, :-for all the subjs
Z matrix is symmetric about the diagonal of NaN.
I am not sure if this is clearer than FAQ, but I hope it will be helpful.
Threaded View
| Title | Author | Date |
|---|---|---|
| Julia Berezutskaya | Aug 3, 2013 | |
| Dan Zhu | Aug 22, 2013 | |
| Julia Berezutskaya | Aug 22, 2013 | |
| Dan Zhu | Aug 22, 2013 | |
| Julia Berezutskaya | Aug 22, 2013 | |