[Mristudio-users] questions about Fiber tracking

[Susumu Mori]

To make our parcellation map overlapped well on your data, you need to use
our LDDMM-based non-linear image alignment. If this cannot align the
parcellation map, then, you need to improve the LDDMM results.

Usually, following steps help;
> use both FA and MD (or trace) map to drive the LDDMM
> segment the ventricles and use them for the third channel of LDDMM


On Sat, Aug 17, 2013 at 4:00 AM, Kotikalapudi Raviteja <
raviteja.kotikalapudi at gmail.com> wrote:

> Dear Susumu,
>
> I have been using JHU-Parcellation atlas for DTI analysis. I want to
> extract the FA maps from corticospinal tract for which I made an ROI from
> the JHU-Parcellation atlas. But it is not overlapping well with the FA
> images. I need to know if there is a way to shift the ROI on the FA image?
>
> Your suggestions would be appreciated.
>
> Many thanks,
> Ravi
>
>
> On Fri, Aug 16, 2013 at 11:03 PM, Susumu Mori <smoriw at gmail.com> wrote:
>
>> Hi Ying,
>>
>> Here is my view of "registration".
>>
>> > Imaging is all about what is where.
>> > When you quantify images, you have to make sure that you are measuring
>> the same structures or locations across subjects. This act of identifying
>> the same thing is "registration".
>> > For example, you can manually define the hippocampus across subjects.
>> By doing so, you identify a group of voxels that belong to the hippo and
>> you "register" the hippo locations across subjects. You defined
>> corresponding voxels across subjects. In this sense, manual ROI delineation
>> is one of the registration methods.
>> > Because the manual delineation is time consuming, people don't want to
>> do it. Then what we do is to align all images and assume all voxels are
>> registered automatically. This means, if I pick up an arbitrary location
>> (like a single voxel), I can immediately locate the corresponding location
>> (voxel) in each subject by using the same x, y, z coordinates.
>> > Tractography can also be considered as a registration tool because, for
>> example, we can identify a group of voxels that belong to the corticospinal
>> tract across subjects. Therefore, there is nothing wrong with defining the
>> cst in each subject without image-alignment.
>> > However, there could be several reasons why you may want to do image
>> alignment (thus voxel-by-voxel registration) first and then additional
>> registration by tractography. First, tractography requires manually (or
>> automatically) defined  ROIs for seed and destination. This could be done
>> more reproducibly if all images are aligned. Second, after tracts are
>> defined, you can easily create population-averaged probabilistic maps by
>> simply adding the 3D masking files that define the tract locations. On the
>> other hand, if you are just interested in voxel count (size of the tract)
>> and FA values, image alignment may not be necessary.
>>
>> Susumu
>>
>>
>> On Fri, Aug 16, 2013 at 2:25 AM, <wyzyf at mail.ustc.edu.cn> wrote:
>>
>>>
>>>
>>>
>>>   Hello professor:
>>>
>>>     As I understood, fiber tracking analysis with DTI studio should be a
>>> group analysis and image registration is
>>>   necessary. Am I right?
>>>
>>>     I have finished the image registration with DiffeoMap through linear
>>> and nonlinear(b0+FA) registration but I don`t
>>>   know how to achieve group analysis and the registrated FA doesn`t
>>> match the origin vector.
>>>
>>>     Thank you very much~!
>>>
>>>                                                     Ying Wang
>>> _______________________________________________
>>> 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
>>
>>
>
> _______________________________________________
> 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/20130817/3d4a85f7/attachment-0001.html