Dear Torben,
surface-based statistics are indeed a feature of SPM12 (briefly
mentioned in the release notes (section 6.6)) and illustrated in the SPM
manual (pp. 432-436) with M/EEG source reconstructed data.
At the moment, input data have to be stored in GIfTI format with two
specificities:
* the name of the file containing the geometry of the mesh has to be
stored in the first metadata entry under the name "SurfaceID".
* the encoding format has to be 'ExternalFileBinary' (dual files,
.gii/.dat).
Here is an example of how the data should be saved:
G = gifti;
G.private.metadata(1).name = 'SurfaceID';
G.private.metadata(1).value = ...
fullfile(spm('Dir'),'canonical','cortex_20484.surf.gii');
G.cdata = randn(20484,1);
save(G, 'data.gii', 'ExternalFileBinary');
spm_mesh_render(G)
These requirements might be relaxed in the future: the mesh geometry is
needed for random field theory, and storing the data as binary files
allows for the reuse of the same memory mapping code used for volumetric
data.
Not all relevant SPM functions have been upgraded to deal with both
surface and volumetric data; the hope is to do so following user needs
when they occur.
Best regards,
Guillaume.
On 22/03/16 15:12, Torben Lund wrote:
> Dear Guilaume and SPM list
>
> In the code of spm_spm, it seems that SPM analysis of gifti surfaces
> should be supported, and indeed this feature seems to be used in the
> newly released CAT12 toolbox, but I have been unable to get it to work
> on gifti files which I have created myself. What are required of gifti
> files if they are to be analysed with spm_spm?
>
> Best
> Torben
>
>
>
> Torben Ellegaard Lund
> Associate Professor, PhD
> Center of Functionally Integrative Neuroscience (CFIN)
> Aarhus University
> Aarhus University Hospital
> Building 10G, 5th floor, room 31
> Noerrebrogade 44
> 8000 Aarhus C
> Denmark
> Phone: +45 7846 4380
> Fax: +45 7846 4400
> http://www.cfin.au.dk
> [log in to unmask] <mailto:[log in to unmask]> (for official stuff)
> [log in to unmask] <mailto:[log in to unmask]> (for faster response)
--
Guillaume Flandin, PhD
Wellcome Trust Centre for Neuroimaging
University College London
12 Queen Square
London WC1N 3BG
|