 |
 |
Dear Sanjay, > Dear FSL experts > I would like to use FSLNETS to create a correlation matrix to compare patients with controls - as per the figure 5 in this neuroimage: clinical paper > > http://www.sciencedirect.com/science/article/pii/S2213158214000631 > > Currently, FSLNETS produces a matrix of full v partial correlation. Is it possible to do the above? Figure 5 of the paper you cited contains two panels: panel A: hierarchy of the average network in the whole sample - this is done with nets_hierarchy (as shown in “nets_examples.m”) panel B: average correlation matrix separately for the two groups and displayed so that one group is above the diagonal and the other is below. To obtain the equivalent of panel B you can: 1) split your netmats1 (assuming you’re using Pearson correlation as in the paper) If you have 20 controls and 25 patients (in this order) you do: netmats1_controls=netmats1(1:20,:); netmats1_patients=netmats1(21:end,:); 2) calculate the mean network for each group [Znet1_con,Mnet1_con]=nets_groupmean(netmats1_controls,0); [Znet1_pat,Mnet1_pat]=nets_groupmean(netmats1_patients,0); 3) display them in one matrix (e.g. controls above diagonal, patients below diagonal) pat_belowdiag=tril(Znet1_pat); con_abovediag=triu(Znet1_con); final_matrix= pat_belowdiag+ con_abovediag; imagesc(final_matrix) > > Also, FSLNETs is not producing a scale for the hierarchy/correlation matrix figure. Is there an easy way to create one? The easiest way is to create the figure, click on the matrix and then click on the “insert colorbar” icon. To include it in your code, type “help colorbar” in matlab command window for details. > > Thanks > > Sanjay Hope that helps, Ludovica — Ludovica Griffanti, PhD Analysis Postdoctoral Research Assistant Oxford Centre for Functional MRI of the Brain (FMRIB) Nuffield Department of Clinical Neurosciences, University of Oxford John Radcliffe Hospital Oxford, OX3 9DU, UK email: [log in to unmask]