Just in case you aren't aware of this, using FA for the registration and then using these registrations to analyze other scalar measures is suggested course of action.  It sounds like you might be asking if EAR could be used to obtain more accurate registration - I imagine the answer is in theory, yes, but that this pipeline currently works best with FA.  Anyone know if any changes would be necessary?  Thresholding might need to be slightly different.

Thomas Ballinger
Research Assistant
PNL at BWH

On Mon, Nov 15, 2010 at 4:49 AM, Saad Jbabdi <[log in to unmask]> wrote:
Hi
You can use repeated and 'sound' calls to fslmaths on the eigenvalues (L1,L2,L3) to calculate this EAR index.

Saad.


On 14 Nov 2010, at 17:14, Arman Eshaghi wrote:

> Dear FSL experts,
>
> Recently ellipsoidal area ratio (EAR) has been introduced as a more sensitive index of white matter integrity than fractional anisotropy(FA). I was wondering about the exact methodology of using EAR instead of FA in TBSS. Is it possible for FSL to calculate images with this index? Should the mean image and its skeleton be created with EAR instead of FA? How valid is using of EAR in TBSS?
> Any input on this issue is gratefully appreciated.
>
> Cheers,
> Arman
>

--
Saad Jbabdi
University of Oxford, FMRIB Centre

JR Hospital, Headington, OX3 9DU, UK
(+44)1865-222466  (fax 717)
www.fmrib.ox.ac.uk/~saad