Hi,
1) Since I am not familiar with the randomise tool from FSL, is it appropriate to convert the 4D skeletonised data and run the statistics in SPM? Is there any differences between randomise in FSL and statistics in SPM (i.e., tstats1 in FSL vs spmT_1 in SPM)?
No, it's not really appropriate for several reasons (such as the non-gaussianity of the TBSS data, see the original TBSS paper for this, Smith et al., 2006), so you cannot use the random field theory as implemented in SPM for your statistics and should prefer non-parametric inference such as in 'randomise'...
2) I would also like to correlate these data with clinical measures only in my patient group. Since the data now is constructed in 4D (including both the patient and control group). Is there a quick way to generate another 4D data for specific subjects or do I have to re-run the analysis and re-generate the mean FA skeleton?
Yes, you can split your 4D image with fslsplit, and then reconcatenate the volumes of interest accordingly with fslmerge.
3) In addition, how do I identify the significant clusters (quantitatively, not visually), and label them according to a given atlas if possible?
With fslview and its atlas tool, please have a look at the manual here: http://fsl.fmrib.ox.ac.uk/fsl/fslview/
Hope this helps,
Gwenaelle
--------------------------------------------------------------------
Gwenaëlle Douaud, PhD
FMRIB Centre, University of Oxford
John Radcliffe Hospital, Headington OX3 9DU Oxford UK
Tel: +44 (0) 1865 222 523 Fax: +44 (0) 1865 222 717
www.fmrib.ox.ac.uk/~douaud
--------------------------------------------------------------------
|