[Mristudio-users] Flip eigen vector

[Miguel Burgaleta]

Dear DTIStudio experts,

My question has to do with the flipping of the eigenvectors in fiber  
tracking. It looks like this is something very tricky, so I would like  
to make sure that I am choosing the right one.

So I ran the fiber tracking ( starting and ending FA thresholds at .2  
and angle at 40) and tried a single 'OR' ROI in the brainstem,  
capturing the corticospinal tract. I used this single ROI to track  
fibers in 4 conditions: no flipping of eigenvectors, flipping of X,  
flipping of Y and flipping of Z.

Please see results in attachment (note that the acquisition is not a  
full brain coverage). I'm not very good at anatomy but would say that  
flipping the Y eigenvector leads to the best results - better than no  
flipping... What do you think?

Thanks in advance for your time.

Miguel
-------------- next part --------------
A non-text attachment was scrubbed...
Name: flipEigenvector.jpg
Type: image/jpeg
Size: 89242 bytes
Desc: not available
Url : http://lists.mristudio.org/pipermail/mristudio-users/attachments/20091211/b1745ca0/attachment-0001.jpg 
-------------- next part --------------