Hi Everybody,
Thanks to some advice that I received from Heather Urry, I have had some success unwarping
my EPI volumes. My question is, is there a way to statistically assess the effectiveness of FUGUE?
I imagine that one could compare the mutual information scores for the registrations of
distorted and corrected volumes to the subjects high resolution anatomical. However, FLIRT
apparently does not report the final values of the cost function. I would appreciate any
suggestions.
Regards,
-Kent
PS. Here is Heather's advice on how to the transform the field map produced by the Siemens Magnetom Trio gre_field_mapping protocol for use by FUGUE:
Hi Kent,
For what it's worth, when I reconstruct our Siemens Magnetom Trio DICOM
fieldmap images using MRIcron, I get 2 magnitude images and one phase map.
The phase map is scaled with slope of 2 and an intercept of -4096. To
convert that to radians per sec, I use the following command:
fslmaths fieldmapping64s006a201.nii.gz -mul 3.14159 -div 4096 -div .00246
fieldmap.nii.gz
fieldmapping64s006a201.nii.gz is the name of the phase map.
Multiplying the phase map by 3.14159 and dividing by 4096 converts the
image to -pi to pi. Dividing again by .00246 converts to
radians/sec. .00246 is the TE difference, in seconds. In my case, TE1 =
2.84 ms and TE2 = 5.3 ms, thus TE diff = 2.46 ms, or .00246 s.
fieldmap.nii.gz is the properly scaled fieldmap that is output by this
command. It's the one that I use for B0 unwarping in FEAT.
Hope this is helpful! Ignore if this isn't what you were looking for. :)
Cheers,
Heather
|