[Mristudio-users] Kimap, eigenvalues, and longitudinal analysis

[Andreia Vaconcellos Faria]

Hi, Jeff,

Q1: Kimap, such as any other matrix or deformation field, will work for any
image that is co registered with those you used for the transformation.
Since all the DTI (eigenvalues, B0, FA, and also tensors) come from the same
file, they are of course coregistered and therefore your Kimap is “valid”
for all of them. The practical advantage of applying Kimap (or any
transformation) to the tensors is that, using just this one-step
transformation, you can calculate various scalar images (eigenvalues, FA…)
from the same “normalized” tensor file. Also, the tensors are "re-oriented"
and there is no lost of the directional information.

Q2: The contrasts that drive LDDMM serve as topological information. The
most common approach is to use one image with large contrast between brain
and “no-brain” (csf and background), such as B0, adc, trace…; and another
channel to provide information about the “internal” structure, such as FA
(that provides contrast within the white matter). The normalization is
improved when you use images that provide different kind of information and,
theoretically, there is no gain on imputing channels that provide the same
kind of information (for example, B0 *AND* trace).
The linear transformation and the LDDMM are independent; you can use B0 in
both steps if you want.

Q3:The idea that drives the imaging process is exactly the same. Of course,
you can choose different contrasts (channels), templates and pipelines
according the design of your study.

I hope it helps.
Best,

Andreia.

On Mon, Jun 28, 2010 at 7:53 PM, Jeff Sadino <jsadino.queens at gmail.com>wrote:

> Hello,
>
> Can someone please verify my understanding of how LDDMM works?  In the
> DiffeoMap download, there are template images for B0, FA, etc., but none for
> the eigenvalues.  Is the Kimap only valid for the images used during the
> LDDMM, or can it be applied to all of the other (scaler) images as well?
> Can it be applied to the eigenvalue images, or would we have to apply it to
> a tensor file, load the tensor file into DTIStudio, and then recalculate the
> eigenvalues?
>
> Second, it seems like the more images/channels that are sent to the LDDMM
> servers for coregistration, the better the Kimap transformation.  Is this
> true?  Since we use the B0 image to drive the linear transformation for our
> FA and Trace maps, which are the values we really are interested in, should
> we also submit the B0 image to the LDDMM transform, or would that
> "over-transform" the Kimap?
>
> Lastly, is there anything different that we should be doing if we are doing
> longitudinal analysis, or is the protocol the same as the cross-sectional
> analysis?
>
> Thank you everyone for your continued help and support!
> Jeff Sadino
>
> _______________________________________________
> 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
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mristudio.org/pipermail/mristudio-users/attachments/20100629/dc8781f2/attachment.html