[Mristudio-users] Fwd: mristudio-users Digest, Vol 148, Issue 3

[Susumu Mori]

On Sat, Aug 31, 2013 at 6:14 AM, Rotem Saar <saar.rotem at gmail.com> wrote:

> Dear Susumu and Ravi,
>
> Thank you for your help ! I think the best option for me is the third one
> - (copying it here)-
>
> 3) As Ravi suggested, you can warp our white matter parcellation map to
>  your subject image and automatically parcellate the corpus callosum into
>  three regions (the genu, body, and splenium of CC).
>
> so my remaining questions (for now) are:
>
> 1) From where can I download the atlas ?
>

You can download RoiEditor or DiffeoMap from www.mristudio.org, which come
with all the atlases.

2) If can can give a short explanation regarding whether I should or should
> not co-register my DTI images to the T1 images ? In what case do we have to
> register the diffusion to the anatomy ? i.e., if I want to extract the
> Corpus-callosum from each subject and compare the FA value between patients
> and controls - Do I have to do this ?
>

There is a typical argument about how to define an ROI to measure  MRI
contrasts. Suppose you are interested in FA of, say, the corticospinal
tract. You get a FA map, identify the CST, and draw an ROI on it.

Then people say, "Wait a minute. It doesn't make sense. You want to measure
FA and draw an ROI directly on the FA map? You can get any value you want.
If you define only the core of the CST, you get a high FA. If you include
the edge, you get a lower FA. It's not reliable."

This leads to an idea of defining an ROI on an unrelated image, like T1,
and then apply it on an FA map. There are several problems, though. First
of all, you can't see where is the CST in the T1. Second, there is always
error in co-registration because each image has a different amount of B0
susceptibility distortion. So alignment of two images with different
contrasts is not easy.

I personally think defining an ROI directly on the map of interest is fine
(like drawing an ROI on an FA map and measure a FA value).
You just have to be careful when you interpret the data.

If a patient has the affected CST, the ROI analysis may return lower FA or
smaller size, depending on how you define the ROI. We can't conclude if the
CST is really smaller or the property (e.g. axonal density, myelination)
changed without size change. However, we can at least detect the change.
Interpretation of MRI is always difficult because, after all, we are simply
detection signals from water.


>
> Thanks again for all your help
> Rotem
>
>
> 2013/8/29 <mristudio-users-request at mristudio.org>
>
>> Send mristudio-users mailing list submissions to
>>         mristudio-users at mristudio.org
>>
>> To subscribe or unsubscribe via the World Wide Web, visit
>>         http://lists.mristudio.org/mailman/listinfo/mristudio-users
>> or, via email, send a message with subject or body 'help' to
>>         mristudio-users-request at mristudio.org
>>
>> You can reach the person managing the list at
>>         mristudio-users-owner at mristudio.org
>>
>> When replying, please edit your Subject line so it is more specific
>> than "Re: Contents of mristudio-users digest..."
>>
>>
>> Today's Topics:
>>
>>    1. Re: FA value for regions in the corpus-callusom (Susumu Mori)
>>    2. Re: FA value for regions in the corpus-callusom
>>       (Kotikalapudi Raviteja)
>>    3. Re: FA value for regions in the corpus-callusom (Susumu Mori)
>>
>>
>> ----------------------------------------------------------------------
>>
>> Message: 1
>> Date: Thu, 29 Aug 2013 11:43:38 -0400
>> From: Susumu Mori <smoriw at gmail.com>
>> Subject: Re: [Mristudio-users] FA value for regions in the
>>         corpus-callusom
>> To: "DTI Studio, ROI Editor, DiffeoMap Questions/Support"
>>         <mristudio-users at mristudio.org>
>> Message-ID:
>>         <CAEq9apJZ+6Dmu3S2bXB0p9CW0rkiLDzQX3GqVU7K9q_S2K=
>> 9UA at mail.gmail.com>
>> Content-Type: text/plain; charset="us-ascii"
>>
>> There are several options depending on what you need;
>>
>> 1) Do fiber tracking. It seems that you obtained 5 different tracts that
>> comprise the corpus callosum. As you mentioned, you can obtain one FA per
>> tract. In your case, 5 FA values. Each FA value is the representative FA
>> of
>> each tract by avaraging FA values of all voxels inside a tract.
>> 2) Instead of calculating a representative FA for each tract by averaging
>> all voxels, you can also use the "profile" option, with which, you can
>> obtain a FA value at each 2D slice (axial, sagittal, coronal separately).
>> 3) As Ravi suggested, you can warp our white matter parcellation map to
>> your subject image and automatically parcellate the corpus callosum into
>> three regions (the genu, body, and splenium of CC).
>>
>> Hope one the approaches addresses your need.
>>
>>
>>
>>
>> On Thu, Aug 29, 2013 at 7:32 AM, Kotikalapudi Raviteja <
>> raviteja.kotikalapudi at gmail.com> wrote:
>>
>> > Dear Rotem,
>> >
>> > Though I couldn't get you completely, I understand that you need to
>> > extract FA values and compare the same between subjects and controls.
>> For
>> > this purpose, you could download the JHU parcellation atlas and then use
>> > MRICro to extract your ROIs using 'apply intensity to volume' option.
>> This
>> > particular ROI would then be saved and you could open it for the FA maps
>> > and note down the values.
>> >
>> > I hope this helps, if it is related.
>> >
>> > Regards,
>> > Ravi
>> >
>> >
>> >
>> > On Thu, Aug 29, 2013 at 4:29 PM, Rotem Saar <saar.rotem at gmail.com>
>> wrote:
>> >
>> >> Dear experts,
>> >>
>> >> I have performed Automatic image registration to 19 subjects (who I
>> have
>> >> DTI scans for) and now about to begin fiber tracking.
>> >>
>> >> According to my previous anatomical results, I found volume reductions
>> in
>> >> specific areas of the corpus callosum, which were correlated to
>> behavioral
>> >> memory results. I now want to analyze the DTI data, and test for the
>> >> correlations between the anatomical volume (per region, 5 in number),
>> >> memory performance and FA values (per region, 5 in number).
>> >>
>> >> In the current working program in DTIstudio I only know how to extract
>> a
>> >> singe FA value for the corpuse if this is what I marked for the fiber
>> >> tracking. I understand that I can mark a single area in the corpus each
>> >> time but that will probably won't be accurate when comparing
>> subjects...
>> >>
>> >> Is there any way to load a segmentation map in order for DTIstudio to
>> >> compute the FA value for each region defined in the map automatically
>> >> without the need to mark each time a different region ? if there is
>> such an
>> >> option, I will really appreciate if u can explain what parameters (file
>> >> type ect...) it should contain.
>> >>
>> >> I really want to come to the tutorial, but the thing is I live in the
>> >> other side of the sea:) so that won't be so easy for me to do...
>> >>
>> >> Thanks again for all your help
>> >>  Rotem
>>
>>
> _______________________________________________
> 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/20130903/d0ad1a0a/attachment-0001.html