Tw-fc


#1

Dear MRtrixers,
I am trying to better understand and replicate the paper about TW-FC (NeuroImage2013)
Specifically I have a question about the FC analysis using an ICA based approach. After doing all the steps using the Melodic software, I’ve applied the “dual regression” obtaining two files for each subject:

  • dr_stage1_subject[#SUB].txt (time-courses file obtained after the spatial regression)

  • dr_stage2_subject[#SUB].nii.gz (FC map obtained from the time courses regression)

But in the (NeuroImage2013) article there is written

The group components were projected back onto each subject’s own data using a ‘dual-regression’ approach (Filippini et al., 2009). Specifically, the group components were spatially-regressed against each individual’s MNI-space data in order to generate subject-specific component time-courses. These time-courses were then partially-correlated against the subject’s own-space data in order to generate a set of connectivity maps for each subject in the same space as their diffusion MRI data.

How can I do this step? I think that for each subject I should only take the “time-courses” file from the first step of dual regression and than partially correlate this file with the subject’s own-space data. Is this correct? If so can you advice me how should I do this partial correlation and what does “subject’s own-space data” mean? Sorry I am really not an expert of FC analysis and I wasn’t able to find such variation of (Filippini et al., 2009) approach.

Thank you in advance.
Simona


#2

Dear expert,
Under the guide of “Generating Track-weighted Functional Connectivity (TW-FC) maps” in “Frequently Asked Questions (FAQ)”, I have been familiar with the method of producing TW-FC map. However, this voxel-wise methods could make the interested regions with multiple fibre populations in the TW-FC map less directly interpretable.My question is how to produce Fibre-specifc TW-fMRI map as demonstrated in Fig. 13 in the article: Fernando Calamante Magn Reson Mater Phy (2017) 30:317–335.