[Mristudio-users] Gradient tables and SPM coregistration

[Dorian P.]

Hi Dr. Mori,

I think I have figured out the pipeline now. Two more questions I have are:

1. I have tested both DTIstudio and ExploreDTI. The second produced
occasionally more voxels with FA over 0.25 following motion and eddy
current correction. As you have recently coauthored a paper with Dr.
Leemans (Diffusion tensor imaging and beyond), I wanted to ask what may be
different to cause different results from the same dataset.

2. You mentioned in another email the possibility to use masks (for
example, an entire hemisphere) during normalization. I haven't found an
option like this in Diffeomap. Is there any such possibility, so that I can
possibly avoid using SPM in the pipeline?

Thank you.
Dorian


2012/12/30 susumu mori <susumu at mri.jhu.edu>

> Hi Dorian,
>
> We don't have experience in mixing SPM transformation and MriStudio
> pipeline. In general, you have to be very careful to mix pipelines from
> different sources;
>
> 1) coordinate systems could be different; right-hand or left-hand
> coordinates / neurology or radiology orientation. You have to be careful so
> that the images and transformation matrices are in correct orders.
> 2) gradient vector re-orientation: If you touch raw diffusion-weighted
> images, you have to transform the gradient vectors too. This is not
> difficult when you use linear transformation, but if you are using
> non-linear transformation, each voxel has to have its own gradient table.
> The procedure becomes simpler if you transform images after tensor
> calculation, but if you transform a tensor field, you need to make sure to
> re-orient the tensor in each voxel.
>
> In general, each software (FSL, SPM, MriStudio) should be already tested
> for the consistency for these issues in a transparent manner, but once you
> mix the pipeline, such a consistency may be lost.
>
> Susumu
>
> On Wed, Dec 26, 2012 at 12:16 PM, Dorian P. <alb.net at gmail.com> wrote:
>
>> Hi everybody, and happy holidays,
>>
>> I am using DTIstudio in a complex workflow.
>>
>> First, I need to make a nonlinear registration using masks on lesioned
>> data. SPM is doing a good job at this point. To apply the transformation
>> matrix produced by SPM, I must export images from DTI studio (analyze
>> format), apply the deformations in SPM and import the files back to
>> DTIstudio. However I am not sure whether a problem arises with the gradient
>> tables not being the same any more. How can I apply the transformation
>> matrix from SPM? Is there maybe a way to convert .mat files to .vtk and do
>> all transofmrations in DTIstudio/Diffeomap?
>>
>>
>> Second, having both T1 and b0 images in two different sessions, which one
>> is better in theory for registration, T1s or B0s ? In many examples the b0
>> has been used for linear (AIR) and nonlinear (LBBDMM) registration. But on
>> the other side a b0 is acquired in a fraction of the time of a T1, so may
>> be less stable to artifacts (or with fat artifacts?).
>>
>> Thank you.
>> Dorian
>> TJU
>>
>> _______________________________________________
>> 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 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/20121231/ddcb86c5/attachment.html