[Mrtrix-discussion] CSD optimal lmax value

[Donald Tournier]

Hi Supreet,

This will depend to some extent on the b-value used - you stated b=0 in
your email, I assume that's a typo. As in the adult case, higher b is
generally better for CSD, even if the images do look much noisier...

In general, the most important step is to get the response function right.
If it looks OK, then the CSD should behave similarly for lmax 6 or 8, with
the only difference really being how sharp the FOD comes out. In my
experience, it doesn't make a great deal of difference which you use in
terms of the tractography, and in neonates, there's no much to be seen in
the DW signal beyond lmax=4... Might be worth pushing to lmax=6 just to
make sure, but I doubt you'd see any improvement going beyond that.
Ultimately, the best thing to do is to try it and see.

For the response function estimation, you're right that with 30 gradient
directions, lmax=6 is the best you can do - thankfully that should be ample
for neonatal data. Again, the best thing is to try it and see what the
response function looks like. If it's well behaved and looks like what you
might expect in a single fibre bundle, then you've probably close enough to
get good results - at least within the limitations of the data you have...

Cheers,
Donald

--
Dr J-Donald Tournier (PhD)

Senior Lecturer, Biomedical Engineering
Division of Imaging Sciences & Biomedical Engineering
King's College London

A: Department of Perinatal Imaging & Health, 1st Floor South Wing, St
Thomas' Hospital, London. SE1 7EH
T: +44 (0)20 7188 7118 ext 53613
W:
http://www.kcl.ac.uk/medicine/research/divisions/imaging/departments/biomedengineering

On 27 May 2014 15:32, "Supreet kaur" <ksupreet6 at gmail.com> wrote:

> Dear MRtrix users,
>
> As I'm new user of MRtrix software. I am processing the neonatal DTI data
> which was acquired with 30 gradient directions and b value of 0 mm/s.
>
> I've used the lmax = 6 for the response function coefficient (since we
> have 30 gradient directions).
>
> For the CSD computation, I've used the lmax of 8 (as its recommended for
> most DWI datasets as per the instructions on the webpage:
> http://www.brain.org.au/software/mrtrix/tractography/preprocess.html
>
> So I was wondering what would be the optimal lmax value for the CSD for
> our neonatal dataset? As our dataset is a 30 gradient DTI not HARDI high
> resolution DWI data.
>
> I'll appreciate your feedback on this.
>
> Thanks in Advance,
>
>
> Sincerely,
>
> Supreet kaur,
> Biomedical research engineer,
> Nationwide Childrens Hospital,
> Columbus, OH
> (614)355-6659.
>
> _______________________________________________
> Mrtrix-discussion mailing list
> Mrtrix-discussion at www.nitrc.org
> http://www.nitrc.org/mailman/listinfo/mrtrix-discussion
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.nitrc.org/pipermail/mrtrix-discussion/attachments/20140528/207046a3/attachment.html>