Hi,
Could anybody help me with the following questions?
Recently I had somebody collect some fMRI data. Unfortunately, the structural
images of most subjects are of only partial brains.
So, the quesitons are as follows:
1. Which way is best for solving this problem?
A. to re-collect structural images. the functional images were collected on last
Sept - does this constitute a problem? Subjects' head positions and directions
for the re-collected structural images will surely largely differ from those for
originally collected functional images - is this a problem? Surely there will be
some practical difficuty for this solution.
B. to preprocess the functional images without structural ones. does it
constitute a serious problem? does it affect its publication seriously? (a good
journal would reject such research result?)
C. any other solution?
2. How to use SPM8 to preprocess fMRI images without structural images?
The routine preprocessing described in "Face fMRI data" in "VII Data sets and
examples" in "SPM8 Manual" needs structural images:
Coregistration and Segmentation use both functional and structural images to
generate a .mat file; Normalisation then utilizes this .mat file.
How can I sidestep this problem? i.e., what should I do in each step?
3. About slice timing and realignment, which one should be done firstly? I
heard slice timing firstly for interleaving images and realigning firstly for
sequencial images. Is it right?
4. I need to do regression analysis (the independent variables in regression are
main independent / manipulated variables in the experiment design). I heard
that it should use the parametric module, right? After implementing the
parametric analysis, there will be no contrast in the contrast window - how to
define the contrast? could you tell me very specially?
Thank you very much!
James
|