NODDI in VBQ Analysis

[hastotes]

Hi,

Just wondering whether you might be able to advise on the following:

I want to do a VBQ style analysis in SPM including MPM (MT, R1, R2s), NODDI (ICVF, ODI), and DTI parameter maps (FA,MD, RD, AD). I've registered my FA images to the R1 images using reg_aladin from niftyreg, and applied the transform to the other diffusion maps. The results look satisfactory.

I have some NODDI error_code maps that contain non-zero voxels. I have removed these voxels from the NODDI output maps by creating a mask based on the error code maps. Most of these voxels at the edges of the brain and so are unlikely to be included in the VBQ white-matter analyses.

However, I have a few participants with non_zero voxels in the white-matter in their error_code maps. Given that I have removed these voxels from the output maps, I am wondering whether you think I should exclude these participants from the VBQ analyses which involve a smoothing step or whether there is anything else you would advise that I do?

Any comments/advice much appreciated.

Many thanks,

HS

[Hui Zhang]

Hi,

There isn't a simple answer to your question.  Having voxels with non-zero error code in the middle of the brain might suggest some data quality issue, e.g. poor SNR.  It is worth to go through your raw DWI to see what is going on.  Should you remove these subjects altogether?  One can argue that this in itself introduces a bias in subject selection.  Ideally, one might want to smooth the data in a way that avoids any influence from the problematic voxels, and avoid just the problematic voxels in the analysis.  This is a more of a question for the SPM forum I think.

Gary

--
You received this message because you are subscribed to the Google Groups "NODDI" group.
To unsubscribe from this group and stop receiving emails from it, send an email to noddi+un...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.