[Mrtrix-discussion] Question about eddy current correction

[Donald Tournier]

Hi Bonnie,

I would agree with your colleague that these data probably don't need
eddy-current correction, for two reasons. First, these data were acquired
on a Siemens Trio, where the only diffusion sequence available is an
eddy-current compensating double spin-echo sequence. This type of sequence
is designed to minimise these problems, and typically brings
misregistration problems down to the subvoxel level. In my experience,
there are no obvious eddy-current artefacts on these types of data.
Obviously, things would be different if the data had been acquired using a
single spin-echo sequence, which is much more prone to these issues.

Second, the approaches available to perform eddy-current correction simply
don't work with such high b-value, low SNR, high contrast images. In my
experience and that of several collaborators, using these methods
introduces far more corruption into the data than was initially present. We
are actively working on methods that would handle these types of data more
reliably, but it will be some time before they are ready for release. In
the meantime, best to steer clear of eddy-current correction methods for
your data...

Hope that helps,

Donald.



On 13 February 2012 17:34, bonnie alexander <bon.alexander at gmail.com> wrote:

> Hi,
>
> I'm returning to an analysis I put aside some time ago and would like to
> check on a step in my protocol I've received differing advice on. My DWI
> data was recorded at Brain Research Institute at the Austin in 2010, and as
> I understand this was using an acquisition optimised for analysis in MRtrix
> (I can provide the details if needed). Could you please clarify whether
> this data requires eddy current correction in e.g. FSL? It's been suggested
> to me by a colleague whose data was acquired similarly that this isn't a
> necessary preprocessing step in this case.
>
> Thanks,
>
> Bonnie.
>
> _______________________________________________
> Mrtrix-discussion mailing list
> Mrtrix-discussion at www.nitrc.org
> http://www.nitrc.org/mailman/listinfo/mrtrix-discussion
>
>


-- 
Jacques-Donald Tournier (PhD)
Brain Research Institute, Melbourne, Australia
Tel: +61 (0)3 9035 7033
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.nitrc.org/pipermail/mrtrix-discussion/attachments/20120222/34bc33b7/attachment.html