I have a couple questions about the NT parameter from the 3dTrackID *.grid output file.
I believe streamline count is another term used synonymously with number of tracts (NT), correct?
Given one subject, if NT = 0 for a region, should all other indices (i.e. FA, MD, RD, etc) also be 0?
Am I correct that NT and PV should be strongly correlated?
Related to the above question, would individual tracts vary in terms of thickness? For example, if NT = 4000 in a region, are all 4000 tracts the same size?
Should we expect any sort of relationship (e.g. positive correlation) between NT and FA within a region?
Q: I believe streamline count is another term used synonymously with number of tracts (NT), correct?
A: Yes, those should be synonymous. NT is the number of tracts in a WM ROI. It is either the number of tracts connecting 2 distinct target ROIs, or the number of tracts going through a single target ROI.
Q: Given one subject, if NT = 0 for a region, should all other indices (i.e. FA, MD, RD, etc) also be 0?
A: Yes.
Q: Am I correct that NT and PV should be strongly correlated?
A: I wouldn’t look for a strong relation here; and it would probably be stronger in DET mode, rather than PROB or MINIP. PV is physical volume, and NT is number of tracts. Tracts are not physiological WM bundles—I prefer to think of them as probability markers for where WM might be running to/from based on subject diffusion data. One could have a large volume with a few tracts found, or a small volume with lots of tracts, etc. There are too many scenarios that could happen to have an expectation of a lasting relationship between a physical volume and estimated tracts.
Q: Related to the above question, would individual tracts vary in terms of thickness? For example, if NT = 4000 in a region, are all 4000 tracts the same size?
A: Tracts are not physiological WM bundles. Each tract is just a line, really without thickness on its own. They provide a path through voxels that might be useful to treat as a probability for associating somehow with physiological WM, but they are really “pointers” of where WM might reasonably be considered to run, based on the diffusion data. Again, I would consider them more as probability markers/streams/lines rather than as representations of WM. As such, they don’t have thickness themselves. They might paint out a WM bundle of voxels that is more or less thick, but that is a separate thing. It is the WM ROI of voxels that they select or highlight that we would most want to work with.
Q: Should we expect any sort of relationship (e.g. positive correlation) between NT and FA within a region?
A: No, I don’t see why there would be any relation between NT and FA.
The
National Institute of Mental Health (NIMH) is part of the National Institutes of
Health (NIH), a component of the U.S. Department of Health and Human
Services.