I've played around with byteswaping, but alas it doesn't seem to be
the issue. The original ANALYZE file is 16-bit little endian (or at
least says it is), but when I swap just the .img to big endian and use
the original header, or give the original .img a 16-bit big-endian
header, the resul
not an easy way. You'd have to find out what the byte swapping is in
each file (or verify that the byteswapping is different). The other
programs may check that the voxel values are "reasonable" and byteswap
them if they are not.
andrewkra...@ucla.edu wrote:
> There's actually quite a large cha
There's actually quite a large chance for that, as we convert the
DICOMs to ANALYZE on iMacs; skull-strip mostly on PCs; register and do
other processing on remote machines running OpenBSD, IRIX, and Linux
(some of which are little-endian and some of which are big-endian);
and then check ev
It looks the same to me in fslview. I'm guessing that there is a
byte-swapping problem. Any chance the .img file was created under linux
and the .hdr created under windows (or vice versa)?
andrewkra...@ucla.edu wrote:
> Sure. Since it's probably a bit big for email, I uploaded it to
> http://dl
Sure. Since it's probably a bit big for email, I uploaded it to
http://dl.dropbox.com/u/432093/ANALYZE.zip
-- Andrew
Quoting Bruce Fischl :
> Can you send us the analyze files?
>
>
>
> On Feb 24, 2011, at 12:37 PM, andrewkra...@ucla.edu wrote:
>
>> I actually didn't know you could load analyze
Can you send us the analyze files?
On Feb 24, 2011, at 12:37 PM, andrewkra...@ucla.edu wrote:
> I actually didn't know you could load analyze into tkmedit, but now that I
> have, it does look like that too (just without the direction arrows). It
> doesn't look like that when I look at it in o
I actually didn't know you could load analyze into tkmedit, but now
that I have, it does look like that too (just without the direction
arrows). It doesn't look like that when I look at it in other viewing
programs, though (it looks like you'd expect).
-- Andrew
Quoting Douglas N Greve :
>
Does it look like that when you load the original analyze into tkmedit?
andrewkra...@ucla.edu wrote:
> Sure. Here's a screenshot of the orig.mgz.
>
> Quoting Bruce Fischl :
>
>> Can you send us an image? The 16 bit shouldn't be a problem
>> Bruce
>>
>>
>>
>> On Feb 23, 2011, at 7:28 PM, andrewkra.
Can you send us an image? The 16 bit shouldn't be a problem
Bruce
On Feb 23, 2011, at 7:28 PM, andrewkra...@ucla.edu wrote:
> The orig.mgz shows up with weird colors (the grey matter is black, the
> background is light gray, and there's a white halo around the brain and in
> the ventricles/CS
The orig.mgz shows up with weird colors (the grey matter is black, the
background is light gray, and there's a white halo around the brain
and in the ventricles/CSF).
Another possibility I just thought of is that the original ANALYZE
image is 16-bit, and is 230 x 230 x 230 voxels. Does mri_c
does the orig.mgz show up properly in tkmedit?
On Wed, 23 Feb 2011, Nick
Schmansky wrote:
> add the flag -notal-check to the end of recon-all command to skip the
> check.
>
> if you can make your average using nifti format you're better off as
> analyze doesnt retain orientation (l/r) info.
>
> y
mri_convert --help has info on specifying orientation, with suitable
warnings about the need for you to make sure you know which way is
correct.
n.
On Wed, 2011-02-23 at 14:55 -0800, andrewkra...@ucla.edu wrote:
> Thanks for the reply. For the orientation, I know from working with
> the sourc
add the flag -notal-check to the end of recon-all command to skip the
check.
if you can make your average using nifti format you're better off as
analyze doesnt retain orientation (l/r) info.
you may have to manually register the brain following this:
http://surfer.nmr.mgh.harvard.edu/fswiki/
Hello,
I'm a complete newbie to FreeSurfer, so I appreciate any and all help
you can provide. I am trying to use FreeSurfer to automatically create
some masks of subcortical structures, but am having trouble getting
the whole process started.
When I run "recon-all -autorecon1 -noskullstrip
14 matches
Mail list logo
