[Mrtrix-discussion] Re: color coding of FODs

[Donald Tournier]

Hi Zhuang,

I'm not sure I completely understand what you're asking. Are you saying you
have a region of fibres running anterior-posterior (green) and some running
inferior-superior (blue), and that in some voxels you get a single FOD peak
with an intermediate orientation/colour? It might help if you could post an
screenshot of the problem (use the MRView 'Orientation plot' sidebar tool,
with the 'overlay' option). If did I understand correctly, then yes, this
does mean that the fibre orientation were not resolved, and yes, this will
influence the tracking results.

Also, in general I wouldn't worry about the colours of the FOD, they're
really only useful for display purposes. The *shape* of the FOD is what
matters, and in your case, a single FOD peak/lobe when you were expecting 2
distinct peaks tells you that the orientations have not been resolved.

Hope that helps.
Cheers,

Donald.


On 14 March 2013 11:07, Zhuang Song <zhuang.song at gmail.com> wrote:

> Hi Donald,
>
> Thank you for addressing my question. It makes perfect sense to color the
> FOD according to the xyz coordinates. My initial question actually mainly
> concerns the symmetry of colors to the center of FOD. As I looked closely
> later at the FOD profiles that MRtrix generated, they were actually
> symmetric to the center. So there is no question for that now. Sorry for my
> earlier misunderstanding.
>
> The region of my interest contains crossing fibers from approximately both
> y (green) and z (blue) directions. If these two different colors were
> mixing in the same lobes of FOD instead of being separated in different
> lobes (as you saw from the example i sent), does it indicate that the FOD
> didn't resolve those fibers from two different directions? In this case, is
> there still any chance for the fiber tracking algorithm to resolve them in
> those voxels? Please let me know if this question is clear.
>
> Thanks,
> Zhuang
>
>
>
> On Tue, Mar 12, 2013 at 5:25 PM, Donald Tournier <d.tournier at brain.org.au>wrote:
>
>> Hi Zhuang,
>>
>> The colour-coding is done strictly according to the X,Y,Z components at
>> that point on the surface, with no regard to which FOD lobe that point
>> might visually appear to belong to. You'll find the full width at half-max
>> of the FOD peaks at lmax 8 are around ~20°, so for an off-axis peak like
>> the one you're showing, the actual peak might be along Z-Y (giving a
>> greenish-blue colour - cyan), but one side of the lobe might be closer to Z
>> (blue) and the other closer to Y (green).
>>
>> Colour-coding according to the closest peak direction could be done, and
>> might be OK when the peaks are well-defined, but there will definitely be
>> cases when two peaks are close to each other and aren't resolved properly,
>> in which case this approach would give misleading results. The current
>> approach is unambiguous and easy to implement, and as far as I know,
>> similar to what other packages might do.
>>
>> Cheers,
>>
>> Donald.
>>
>>
>>
>> On 13 March 2013 10:47, Zhuang Song <zhuang.song at gmail.com> wrote:
>>
>>>
>>> I am having some problem to understand the color coding of the FOD
>>> profile constructed by the CSD method of MRtrix. For my understanding, the
>>> color should be the same along the same axis of the FOD. However quite
>>> often I see the color changes a lot along the same orientation of FODs.
>>>
>>> I attached one example picture of FOD here but not sure if you can see
>>> the attachment in this maillist. In this example, the FOD color changes
>>> from green to blue in the first primary orientation. What should I
>>> interpret it? Does it mean that the primary direction is a mix of diffusion
>>> in both green and blue directions?
>>>
>>> Thanks,
>>> Zhuang
>>>
>>
>>
>>
>> --
>> *Dr Jacques-Donald Tournier
>> *
>> Research Fellow
>>
>> The Florey Institute of Neuroscience and Mental Health
>> Melbourne Brain Centre - Austin Campus
>> 245 Burgundy Street
>> Heidelberg  Vic  3084
>> Ph:  +61 3 9035 7033
>> Fax:  +61 3 9035 7307
>> www.florey.edu.au
>>
>>
>


-- 
*Dr Jacques-Donald Tournier
*
Research Fellow

The Florey Institute of Neuroscience and Mental Health
Melbourne Brain Centre - Austin Campus
245 Burgundy Street
Heidelberg  Vic  3084
Ph:  +61 3 9035 7033
Fax:  +61 3 9035 7307
www.florey.edu.au
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.nitrc.org/pipermail/mrtrix-discussion/attachments/20130318/6636ec13/attachment.html