If you coregister the B0 and fMRI data, how well aligned are they?
Perhaps both datasets have similar patterns of distortion, in which case
the images may match reasonably well. You can try this with the Coreg
and CheckReg buttons.
Best regards,
-John
-----Original Message-----
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]]
On Behalf Of Mahinda Yogarajah
Sent: Thursday, June 28, 2007 4:14 PM
To: [log in to unmask]
Subject: [SPM] Diffusion and fMRI
Dear John and SPM list users,
I am trying to use fMRI activation (eg verbal fluency task) as a seed
point
for DTI tractography in native space. Because the acquisitions are not
distortion matched I need a method of taking a native fMRI activation
blob
and placing it in diffusion space, in order to use this as a start
region.
The method I had in mind was as follows
1) Normalizing (with non-linear warping turned off) a subject's high res
EPI
image (which we acquire with our fMRI images and is co-registered to the
mean fMRI in each subject) to that subject's B0 diffusion image.
2) Then apply that transformation matrix to the blob before reslicing
with
respect to the b0 image.
3) I should also point out that we are applying the fsl eddy correct
algorithm to our diffusion images for EC correction and some motion
correction before using the diffusion images in the above two steps.
Does this seem like a reasonable method, or are there more optimal
methods
that people might suggest ? I have little experience of the diffusion
toolbok or its motion correction tools that I have read about on the
list -
does anyone think that this would be better ?
Thanks.
Dr Mahinda Yogarajah
CRF
DCCE ION
|