Hi, I’m new on MRtrix3. My question is simple. Reading the documentation I know there are 4 types of different options in order to have a structural connectome by tck2connectome command:
- -scale_file path
Why I should use an option instead of another? The choice should be based on what? Each type of these options is suggested in which type of cases?
Thanks in advance
I think it’s better to think not about “what options are available in the command”, but about what kind of contrast you want your generated connectomes to contain, and then how to obtain that contrast using the provided functionality.
-scale_invnodevol are implementations of heuristic corrections that are there for feature-completeness, but I regularly advocate against using.
-scale_length is principally for extracting the mean streamline length of each connectome edge (when used in conjunction with
-scale_file provides the most general case of “I’ve determined through some other process how much each streamline should contribute to the connectome; now I just want to import that data”. The typical use case is sampling some quantitative image along each streamline trajectory (e.g. using
tcksample), and then calculating the mean of these values across all streamlines within each edge. But this has intentionally been made more general compared to previous implementations in order to maximise the possibilities of what could be done with it.
Specifically regarding the -scale_invlength option, please refer to the discussion in the following paper (co-authored by Rob, by the way )