I recommend trying MRICron's dcm2nii(gui) which will take care of the slice
angulation and generate NIFTI files and bvecs that don't require further
modification.
Peace,
Matt.
-----Original Message-----
From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf
Of Chris Koller
Sent: Tuesday, October 13, 2009 7:44 AM
To: [log in to unmask]
Subject: [FSL] Gradient Directions for Philips
Hi
Searching though the archive that there are many postings regarding gradient
tables, however I cannot find a satisfactory solution to my problem. So
apologies for covering 'old ground' again!
I have just started DTI on a Philips 1.5T scanner. Software release
11.1.4.5
and I am not clear on the correct bVecs to use. I am post processing using
FSL and understand that DTIfit requires the bVecs to be in the image co-
ordinates.
I am using the following protocol:
Transverse images
foldover RL
fat shift L
Gradient Resolution Medium
Gradient Overplus No
Using the applet from
http://godzilla.kennedykrieger.org/~jfarrell/OTHERphilips/GUI.html I can
generate the gradient directions, and note that it is independent of slice
angulation.
Examining the dicom tags on the DWI images I also get information about the
gradient directions. With no angulation the dicom tags agree with your
applet. However when the slices are angled the tags have different values
to
your applet. I expect this difference is due to the co-ordinate frame used.
The problem is I'm not sure which co-ordinate frame is which.
Looking at other web sites it appears that, using the above settings, the
dicom tags are in the image co-ordinates. I gather that the applet
generates
in the scanner co-ordinates. Does this mean that the scanner applies
gradients in the same directions irrespective of the angulation, or does it
reorientation for the slice angulation?
Can anyone help/ ? And is there a quick easy way to extract the gradient
directions? I am unable to obtain par/rec files from this scanner.
Thanks in advance
Chris
|