Hi,
Yes, it seems that a minority of machines are experiencing
problems with Analyze files and non-zero origins, probably
due to some library differences with some of our C calls (as
the C++ code doesn't seem to be affected).
Our recommended solution is to use the nifti format.
If you must have Analyze to pass onto other programs, then
convert the nifti format to Analyze after all the FSL processing
is done by using avwchfiletype. This solution has worked for
other users and so hopefully this will solve your problems.
All the best,
Mark
On 24 May 2006, at 18:39, Jeffrey Spielberg wrote:
> Hi, we have discovered that using avwmaths produces this change in
> "mm" values as well. Is this something that we need to worry about
> happening in Feat as well? We have also found that using bet2 does
> not produce the same weird change in "mm" values, but it does set
> the x value for pixel size to a negative value ( i.e. pixdim1 =
> -3.4375 where before it had been positive). Will this produce a
> problem later in Feat? Thanks,
> Jeff
>
> On 5/16/06, Mark Jenkinson < [log in to unmask]> wrote:Dear Jeff,
>
> I'm having a lot of trouble reading your open office format.
> Could you please resend as plain text.
>
> All the best,
> Mark
>
>
> On 16 May 2006, at 20:09, Jeffrey Spielberg wrote:
>
> > Hi, we are also having this problem after using bet. I am
> attaching a
> > document with the details for our system that you asked Matt for, in
> > the hope that this can help you track down the problem, thanks,
> > Jeff
> >
> > On 5/15/06, Mark Jenkinson < [log in to unmask]> wrote:
> >> Dear Matt,
> >>
> >> I cannot replicate this problem at our end with any combination
> >> of Analyze or Nifti.
> >>
> >> Could you please let me know the following details:
> >> - your platform
> >> - the version of FSL you are using
> >> - the avwhd output of the input and output images
> >> - the exact commands that you are running
> >>
> >> Hopefully this will let us track it down.
> >> All the best,
> >> Mark
> >>
> >>
> >>
> >> On 15 May 2006, at 19:31, Matt Holloway wrote:
> >>
> >> > Hi,
> >> >
> >> > I'm trying to brain extract and noise-reduce a set of T1
> structural
> >> > images
> >> > to get these images ready for VBM analysis in SPM2. While the
> >> > process of
> >> > Brain extracting and SUSAN are going smoothly, and nicely brain-
> >> > extracted
> >> > and de-noised images are being produced, I can't figure where the
> >> > coordinate
> >> > system is getting modified. Immediately following Brain-
> Extraction
> >> > the "mm"
> >> > values in FSLview on the X, Y, and Z planes are changing from the
> >> > original
> >> > ~-90mm to +90mm on say the X-axis, to something like 21090mm to
> >> > 21170mm.
> >> >
> >> > I believe this modification of the raw-coordinates may be playing
> >> > into so
> >> > difficulties making a custom t1-template from these images
> down the
> >> > road.
> >> > I'm also seeing some changes in the header files evident by using
> >> > avwinfo,
> >> > in terms of the size of each dimension.
> >> >
> >> > Any insight you could offer would be wonderful.
> >> >
> >> > Thank you
> >>
> >> <bet_details>
>
|