Yes, Matt is right. In fact, you could say that the waytotal reflects trackability differences - for physical and (patho-)physiological reasons.
Cheers,
Andreas
________________________________________
Von: FSL - FMRIB's Software Library [[log in to unmask]] im Auftrag von Matt Glasser [[log in to unmask]]
Gesendet: Freitag, 21. Januar 2011 00:10
An: [log in to unmask]
Betreff: Re: [FSL] AW: [FSL] waytotal vs FA
I don't think waytotal values are comparable across subjects for
trackability difference reasons. You could maybe use it if you normalized
with a "control" pathway you did not expect to see differences in between
groups, however there are probably still reasons this would not be the best
measure. FA is definitely not a perfect measure, but I haven't seen
anything come along that is better (beyond perhaps thinks like radial and
axial diffusivity).
Peace,
Matt.
-----Original Message-----
From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf
Of Andreas Bartsch
Sent: Thursday, January 20, 2011 4:25 PM
To: [log in to unmask]
Subject: [FSL] AW: [FSL] waytotal vs FA
Hi,
well - this is more complicated. The waytotal number is NOT a
straightforward measure of tract integrity. It simply counts how many
samples make it from seed to target. Now you can imagine a scenario where a
tract is "bundled" to a high FA. However, from all the samples you send out
only relatively few go with the "bundle" and make it to the seed. On the
other hand, you can have a connection that fans out, possibly in an area
with lots of xing fibers, but still - lots of samples make it to the seed
along widely different routes but FA remains low. In addition, the closer
seed and target are the more samples will make it between the two - even at
low FA values.
hth,
Andreas
________________________________
Von: FSL - FMRIB's Software Library [[log in to unmask]] im Auftrag von Alan
Plotzker [[log in to unmask]]
Gesendet: Donnerstag, 20. Januar 2011 21:42
An: [log in to unmask]
Betreff: [FSL] waytotal vs FA
Hi everyone,
I'm working on a study examining the relationship between a behavioral
measure and a particular white matter pathway defined using probtrackx (part
of the uncinate fasciculus), and I'm trying to make sense of some of the
results. Specifically:
1. The behavioral measure is negatively correlated with the waytotal for the
left hemisphere pathway but positively correlated with average FA along the
pathway.
2. There is a strong positive correlation between waytotal and average FA
along the pathway in the left hemisphere, but no correlation in the right
hemisphere.
The second point in particular seems strange because there's no reason to
expect major hemispheric differences in the pathway.
I wanted do see if anyone could explain how exactly waytotal is normally
related to FA. I know that the probtrackx algorithm doesn't directly use FA
unless that option is specified, but it seems like the waytotal and FA would
be influenced by similar factors (e.g. more strongly directional diffusion
would likely lead to both higher FA and more high-probability pathways that
make it to the target region). Am I missing anything here?
I appreciate any thoughts or input that you might have.
Thanks,
Alan
|