The only thing -cifti-convert generally tries to preserve is the data
matrix, all the header info is not retained, and to convert it back, most
formats require a template cifti file that has the appropriate header
info.  You can get the relevant header info with
the -cifti-export-dense-mapping command, or dump the XML from the header
with -nifti-information -print-xml .

For using dconns in matlab, you might want to try the gifti toolbox and
ciftiopen from
https://wiki.humanconnectome.org/display/PublicData/HCP+Users+FAQ , as the
files shouldn't be quite as large as csv.  The MEG-derived code currently
will expand the dconn when it is loaded into matlab by adding rows and
columns, so it would not be my first choice currently.

The HCP cifti files have information about 2 surfaces and lots of
voxel-based structures (dconns have this on both dimensions), if you want
to deal with only one surface at a time, the -cifti-separate command may be
what you want.  However, it is a little tricky to separate along both
directions of a dconn file to get, for instance, only the left surface
connectivity to itself.  Could you give us a little more detail as to what
you want to do with the data?  If we knew the specifics, we might be able
to come up with an easier way to do what you want.

Tim


On Fri, Nov 13, 2015 at 1:49 PM, Rose Tharail John <[email protected]>
wrote:

> Hi all,
>
> I am attempting to use the Group average functional connectivity matrix
> (“dense” functional connectome, the grayordinate × grayordinate full
> correlation matrix). Firstly I must convert it from the .nii format to the
> csv. I tried the MATLAB way, but could not succeed. So I decided to use
> workbench command - *wb_command -cifti-convert to -to-text* to convert
> the file into a csv. That worked well producing a csv file. But now, I am
> wondering what all the columns and rows refer to. The conversion to text
> file has removed any information about headers and therefore now I have a
> huge file which I can not make much sense of.
>
> Could anyone please help me figure this out? Also, my end goal is to
> extract only the huge connectivity matrix with the information about the
> surface vertices.
>
> Any pointers will be highly appreciated.
>
> Thanks everyone.
>
> Rose.
>
> _______________________________________________
> HCP-Users mailing list
> [email protected]
> http://lists.humanconnectome.org/mailman/listinfo/hcp-users
>

_______________________________________________
HCP-Users mailing list
[email protected]
http://lists.humanconnectome.org/mailman/listinfo/hcp-users

Reply via email to