Try something along these lines:
caret_command -metric-math-postfix $OUTMETRIC $OUTMETRIC "$REGION sum"
"@$OUTMETRIC:$COLUMN_NAME@ @$OUTMETRIC:$REGION sum@ +"
________________________________
From: Colin Reveley <[email protected]>
To: [email protected]
Sent: Sat, May 26, 2012 1:09:34 PM
Subject: [caret-users] metric sum
Hello.
If I have a surface with some paints on it and a metric on it, then I can
select
the nodes where the paint is "P" and within "P" the metric is non-zero.
and I can do all kinds of stuff subsequently with this subset of nodes (metric
present in region).
I can find the mean, std, etc. I can even do ANOVAs and one or two stats that I
don't even know what they are.
but I want to do is simply add up all the metric values in the selected nodes.
Can this be be done? An option ultimately might be to save the selected to a
new
metric file, and load that into matlab and sum them. which might go wrong since
I only half know what I'm doing there. So, can the sum be computed with caret?
I want to compare the total (sum) amount of metric present in one anatomical
region to the total in another. the metric is the number of "probabilistic"
streamlines that terminate in the GM of the regions, and hence the sum gives me
an estimate (actually it should be the same value, in principle) of how many
lines have hit each tracking classification mask made from the paint regions.
maybe there's a better, more elegant way to estimate that information than
summing the metric, given that the metric is necessarily mapped to the surface
from volume imperfectly?
best,
Colin
_______________________________________________
caret-users mailing list
[email protected]
http://brainvis.wustl.edu/mailman/listinfo/caret-users
