geodaisics / tractlearn-wholebrain Goto Github PK

Hi @ArnaudAttye,

I have few questions on the processing pipeline outlined here:

• Should the population template be computed from "control only" or from "control + patients"? The page suggests

  As TractLearn requires an excellent matching between subjects, the first steps imply non-linear coregistration in a common template space. Here we assume that you have already created a template using for example population_template coming from MRtrix (https://www.mrtrix.org/).

• Should the DWI volumes be all registered to MNI space? For poplation-template and TractSeg outputs? Or is it only required for FA volumes? It suggests

  FA analysis does firstly require to coregister all individual FA maps (named FA_MNI.nii.gz as it is assumed to be in the MNI space for TractSeg)

• If we want to add other parameters to compute z-score, MD for example, should the pathway be similar as used for FA?

  Note that this framework can easily be extended to other metrics.

Thanks a lot for your time. It is a great tool, having fun using it :)