Dear Moran,
The DTI model effectively estimates a 3D ellipsoid, which is fully described
by its three orthogonal axes. The lengths of these axes are proportional to
the tensor eigenvalues, while their orientation is given by the three tensor
eigenvectors.
As you correctly said, the principal eigenvector (the one associated with
the maximum eigenvalue), gives the orientation of the largest axis of this
ellipsoid and provides an estimate of the dominant fibre orientation in a
voxel. The other two eigenvectors have no direct biophysical meaning, they
are perpendicular to the principal one and all of them define a local
coordinate system for the particular voxel.
There is no calculation involved in the RGB colour coding, so no extract
image to extract. It is just a different way to display the principal
eigenvector image. If a modulation is chosen, e.g. using the FA, then the
x,y,z components of the principal eigenvector are multiplied in each voxel
by the FA value.
Cheers,
Stam
----- Original Message -----
From: "Moran Artzi" <[log in to unmask]>
To: <[log in to unmask]>
Sent: Thursday, November 03, 2011 7:25 AM
Subject: [FSL] dtifit output
Dear experts,
I noticed that the each one of the eigenvector maps – the output of the
dtifit code, consist of 3 images
I assume that the first one in the principle direction but what represent
the second and third map?
In addition is it possible to extract into image the final calculated vector
(as displayed in fslview using RGB coding)?
Thanks a lot
Moran
|