Hi Saad,
ok, creating connectivity matrix is not that easy as I though, seems that I am one of the rare person to use surface and trying to understand the output. =-/
So as ROIs are defined on vertices and not triangles I tried to divide any fiber count (matrix3 entry) with endpoint in surface by 3, thus by 9 when the fiber has both ends in surfaces. Still there is a factor of 8 to the theoretic max number of fibers tracked...
Also just to be clearer: the surface files are gifti as I am manipulating HCP data, but I created these gifti to resemble what is outputted by label2surf, which is 3 dataarrays : coordinates, triangles and label (used as binary mask by probtrackx2). Gifti files carries an affine transformation, thus probtrackx seems not to need a reference volume for the surface if I am right? But it if really was the problem the results would look weird but this is not the case.
Sorry for bothering so much, I am just worried about doing something wrong being a newbie to fdt.
Cheers.
basile
|