[Mristudio-users] Late Stage AIR?

[susumu]

Hi Dylan,

We use AIR-based image co-registration for two purposes;

1) Within one DTI dataset, the images (e.g. b0 and 30 diffusion-weighted
images) may not be perfectly registered because of subject motion. We do
this type of "intra-DTI data" registration to improve the accuracy of tensor
calculation. This has nothing to do with the orientation of the brain such
as AC-PC alignment.

2) If you get DTI data from multiple subjects, their brain orientations may
not be well aligned. In this case, if you try to extract a "corresponding
axial slice" from all subjects, you may have a hard time. This may affect
reproducible fiber tracking because it relies on visually finding equivalent
2D slices to draw ROIs.

Because you mentioned "Our subjects were reasonably aligned", I assume that
you are talking about #2.

Question 1) It is preferable that the brain orientations are aligned to
define seed ROIs reproducibly. We usually scan subjects in AC-PC aligned. On
the other hand, reconstruction of some tracts is relatively insensitive to
the exact locations of see ROIs if multiple ROIs are used to impose
anatomical constraints. So, it depends, but it doesn't hurt to align the
brain at scanning.

Question 2) Landmarker can re-orient the brains. There are, theoretically,
several ways to do this. For example, you can re-orient raw DWIs and then do
tensor calculation. Alternatively, you can finish tensor calculation and
then re-orient the tensor. If you are using Landmarker, it is much easier to
do the second way. Once the tensor is re-oriented, you can recalculate FA
and eigenvectors. In this way, because you already finished tensor
calculation once, you don't have to do re-flagging. 

Question 3) there is no good answer to it. We usually do AIR for intra-DTI
alignment and then do quality check. Again, once you get a tensor file in
the original space, you can re-orient the tensor to the standard AC-PC
orientation using Landmarker.

Susumu

-----Original Message-----
From: mristudio-users-bounces at mristudio.org
[mailto:mristudio-users-bounces at mristudio.org] On Behalf Of Dylan Alegria
Sent: Thursday, May 28, 2009 3:18 PM
To: DTI Studio, ROI Editor, Landmarker Questions/Support
Subject: [Mristudio-users] Late Stage AIR?

Hi,
I'm still a novice with DTIStudio, and I've run into a concerning  
issue during fiber-tracking.  We collected a 30 direction DTI sequence  
with 2 repetitions on 25 subjects and have run it through DTI Studio  
to do fiber tracking.  We did not initially register with AIR because  
we believed that our subjects were reasonably aligned and we didn't  
want to add artifacts from registration if we didn't have to.

While drawing ROIs for fiber tracking some differences in tilting are  
obvious.  We are considering running AIR on the processed data to  
remedy this problem.  This brings me to 3 questions.

1) Given that fiber tracking operates in a 3D space, how important is  
it to have subjects in exactly the same alignment for drawing the seed  
ROIs?
2) Is it possible to register the FA/Eigenvector/etc. images without  
recalculating them, or going back and re-flagging bad slices?
3) Is it of vital importance to run AIR before doing the quality  
checks for some other reason than having similarly aligned subjects?

Any help would be greatly appreciated.  Thank you very much in advance!

Dylan


Dylan Alegria
CIBSR Imaging RA
Stanford University
dalegria at stanford.edu
(650) 736 0732

_______________________________________________
Mristudio-users mailing list
Mristudio-users at mristudio.org
http://lists.mristudio.org/mailman/listinfo/
Unsubscribe, send a blank email to:
Mristudio-users-unsubscribe at mristudio.org mristudio-users