Originally posted by Alfonso Nieto-Castanon:

Hi Calvin,

Right, those extra columns represent the connections with additional ROIs in your project (beyond those included in your first-level RRC analysis), and they are simply disregarded by most analysis procedures in CONN. In particular the "names" variable represents the names of the ROIs in the rows of the "Z" matrix, and the "names2" variable represents the names of the ROIs in the columns of the "Z" matrix (and "names2" is organized to have its first entries always match those of "names"). That means that when working with these matrices manually you can often simply disregard those additional columns, e.g. taking the first 164 columns of your 164x167 matrix will produce the expected square matrix between pairs of ROIs included in your first-level analysis, e.g. 

  Z_new = Z(:,1:size(Z,1));   % square connectivity matrix

  names_new = names;       % ROI names in rows&columns of Z_new

Hope this helps

Alfonso

Originally posted by Calvin Hew:

Sorry if this is obvious but I am still new to using matlab/CONN. I'm having issues working out where my pipeline has gone wrong.

Just using the CONN toolbox, I've done preprocessing, denoising and first level analysis and from what I understand the .mat file produced at the end of first level analysis provides the matrix representing a functional connectivity network: resultsROI_Subject001_Condition001.mat

The issue I'm having is: the dimensions of names and names2 (representing the dimensions?) are 164x167. From what I know, they should be the same? e.g. 167x167

I've been running through Andy's brainbook tutorials following the default setup esentially just to get an idea of what a network should look like. But I am not sure if anyone may know the cause of this issue or even where we may specify these dimensions?

 

Ah thank you so much! That makes sense now - I've been able to append those new fields on to the corresponding .mat files now. I really appreciate the explanation!