[Mrtrix-discussion] Variability in diffusion acquisition gain affecting FA measures

[shawna farquharson]

Hi Amanda,


I just thought I'd add a quick note to Rob's earlier comment.
Unfortunately, (or fortunately) the issue with pre-scan normalise
'randomly' turning off on the Skyra systems is not entirely random. Below
are a few notes to describe our experience to date:

   - We have noticed this problem occurs when the 'virtual coils' do not
   match the 'actual coils' at the time the sequences are bought across from
   the protocol tree. This happens if the coil is open or if only the base
   section is mounted, and is regardless of virtual coils set.
   - If this occurs, the scanner will display a prompt to warn it is
   changing to the body coil and it will turn off pre-scan normalise on all
   subsequent sequences.
   - The coils & pre-scan normalise option then has to be set for each
   sequence individually.
   - There is also an issue with the 'copy everything' option when copying
   sequence parameters because pre-scan normalise and auto coil select do not
   copy in this instance.
   - We now wait until the subject is in the scanner and the coil to be
   used is closed before we move sequences from the protocol tree to avoid
   these issues.

We do not experience any of these problems with our Trio, so I think it may
be a glitch in the programming. Siemens are aware, however, I sure they
will apprcieate more than one site bringing this to there attention.

I hope this information is useful,
Shawna

-- 
-----
Shawna Farquharson
Chief Research Radiographer
Brain Research Institute,
Melbourne Brain Centre,
245 Burgundy Street,
Heidelberg,
Melbourne 3084
 -----
Phone: (03) 9035-7139
Email: s.farquharson at brain.org.au
-----



---------- Forwarded message ----------
From: Amanda Ng <amanda.ng at monash.edu>
Date: Fri, Jan 18, 2013 at 6:40 PM
Subject: Re: [Mrtrix-discussion] Variability in diffusion acquisition gain
affecting FA measures
To: Robert Smith <r.smith at brain.org.au>


Thanks Robert! I think we might've just figured it out. Turns out that
the Skyra was 'randomly' turning off the pre-scan normalise that we'd
set as default. We're going to test it next week, but pretty sure this
is what caused the loss in gain.

Ta!
Amanda

On 18 January 2013 17:45, Robert Smith <r.smith at brain.org.au> wrote:
> Amanda
>
> Just in case nobody else has chimed in yet, a couple of ideas of what to
> look for in the DICOM headers:
> * Head coil: Do you have more than 1 head coil for your Skyra? Should be
> easy to find the number of channels within the headers, it's entirely
> possible that a radiographer may have fitted the wrong one for those
> particular scans.
> * FFT scale factor: The receiver gain is either high or low (20dB
difference
> from memory) so wouldn't cause what you're seeing, but the FFT scale
factor
> scales the raw A/D converter data before (?) application of the FFT. Not
> sure how this could be changed in the protocol without you knowing about
it,
> but I've seen the Siemens console software do weirder things.
>
> Also just in case: if the images with lower mean signal actually appear to
> have higher SNR, that would suggest the coil combine mode was changed.
>
> Rob


On 16 January 2013 17:44, Amanda Ng <amanda.ng at monash.edu> wrote:

> Hi all,
>
> I was hoping someone might be able to explain an 'anomaly' I've come
> across.
>
> We have some 64 direction diffusion scans taken on our Siemens Skyra
> (10 subjects, 2 time points each). We've used the same protocol for
> all the scans. Looking at the DWI files (ie b0 volume + 64 directional
> volumes), most of the b0 volumes have a mean intensity of ~160-210 and
> the directional volumes have mean intensities of ~18. There are three
> scans that have mean B0 intensities of ~70 and mean directional volume
> intensities of ~6. This drop in intensity is affecting the FA measures
> (we assume this is because of the lower SNR). I've ruled out software
> and hardware upgrades as a cause. I've been told there were no changes
> to the protocol, and I can't find anything in the DICOM headers
> indicating a change in protocol.
>
> Would anyone know why this is occurring and how to avoid it? Has
> anyone come across it before?
>
> Thanks in advance!
> Amanda Ng
> _______________________________________________
> 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/20130131/895db568/attachment.html