Andy:
I have not used ImageJ, but NIH Image, the non-java version, works on
only 8 bit image data. This will greatly compromise your input data,
as the signal resolution (the detectable signal difference in each
image) will be about 0.4%. Make sure that this is not the case with
ImageJ. If it is, use different software for conversion (you can find
some on my site at www.brainmapping.org, as well as many other
places) that leaves the full 16 bit resolution.
'to3d' takes care of the image orientation because you help it to
define the orientation field in the analyze headers. It preserves 16
bit resolution.
If you look at:
http://airto.loni.ucla.edu/BMCweb/HowTo/FileFormats.html
you will find descriptions of the different header/data formats in
common use (at least in our place). Note that the analyze format,
used by fsl, can be seen as equivalent to the original Signa data,
less the binary header that precedes each image, coupled to a small
'analyze' header that describes the image contents. There are many
public domain programs with which you can create such headers. I have
one at:
http://airto.loni.ucla.edu/BMCweb/HowTo/EditAnalyzeHeader.html.
--
Mark S. Cohen, Ph.D.
UCLA Brain Mapping Center
660 Charles Young Dr. S.
Los Angeles, CA 90095
PLEASE NOTE:
office: 310/980-7453
lab: 310/825-9142
http://www.brainmapping.org/
We can always use MRI volunteers. If you are interested, go
to http://www.brainmapping.org/ and select 'MRI Volunteers Needed'
|