I think to use randomise you will have to have the subject dimension as the 4th, and hence will need to reorder the data dimensions (e.g. in matlab) first.  This would be straightforward if you were wanting to do voxelwise stats in randomise, but not so straightforward if you want to do a space+time TFCE-like thing. 

I guess what you might do that would be closeish would be to take a 4D space*time dataset from a given subject and feed it into fslmaths, first doing a -bptf temporal lowpass (smoothing) followed by 3D TFCE enhancement (which I think will work separately on each timepoint) and save the 4D output. Then feed that into matlab and reorder the dims so that two spatial dimensions get unwrapped into one, and move time into the 3rd dimension.  Then you can combine across subjects and feed into voxelwise randomise…….how does that sound?

Cheers.


On 11 Dec 2010, at 23:58, Laurence Hunt wrote:

Modelling over subjects, so have 4D (space*time) outputs.

L

On 11 Dec 2010, at 17:37, Stephen Smith wrote:

Hi - over what dimension are you modelling/permuting?  I'm not quite sure what you're saying you need….
Cheers.


On 10 Dec 2010, at 19:04, Laurence Hunt wrote:

Hello,

Is there any way of getting cluster/randomise to produce 4D clusters? I have 5D input data for my higher level MEG analysis (3 spatial dimensions, 1 time dimension, 1 subject dimension). I am currently splitting across time, running randomise at each timepoint, and recombining timeseries, but it would be a lot more powerful if I could apply TFCE or cluster-based thresholding in 4D, rather than 3.

If this isn't possible in randomise at present, is there any way to get cluster to process 4D input images? At the moment it seems to just process the first slice. It would be very useful if I could identify which clusters are contiguous in time...

Thought I'd post to the list in case anyone else out there is interested in doing this :)

Cheers,
Laurence

===========================================
Laurence Hunt, DPhil Student
Centre for Functional MRI of the Brain (FMRIB),
University of Oxford
[log in to unmask]
Phone: (+44)1865-(2)22738
===========================================



---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director,  Oxford University FMRIB Centre

FMRIB, JR Hospital, Headington, Oxford  OX3 9DU, UK
+44 (0) 1865 222726  (fax 222717)
[log in to unmask]    http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------






---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director,  Oxford University FMRIB Centre

FMRIB, JR Hospital, Headington, Oxford  OX3 9DU, UK
+44 (0) 1865 222726  (fax 222717)
[log in to unmask]    http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------