[pymvpa] Fwd: Surface based pattern analysis problems

[Xixi Wang]

Hi Nick,

Thank you very much for your reply!

On Wed, Feb 25, 2015 at 4:52 AM, Nick Oosterhof <nikolaas.oosterhof at unitn.it
> wrote:

> On 24 February 2015 at 21:06, Xixi Wang <wangxixi577 at gmail.com> wrote:
>
>> 5) *My question is: For each voxel of interest inside of the VT cortex,
>> I'd like to get the corresponding surface coordinates. My impression is
>> that SUMA is used for visualization only, so is there any method that I
>> could use to extract certain information? *
>>
>
> See mvpa2.misc.surfing.volgeom, and the ijk2xyz and xyz2ijk methods. For
> example:
>
> from mvpa2.suite import *
> ds=fmri_dataset('my_nifti_file.nii')
>
> # instantiate volgeom instance from fmri dataset
> vg=volgeom.from_any('my_nifti_file.nii')   # can also take 'ds' directly
>
> # convert voxel indices in 'ds' to 3D coordinates
> xyz=vg.ijk2xyz(ds.fa.voxel_indices)
>
> I went through the methods you mentioned and understand that the ijk2xyz
command can convert sub voxel coordinates/indices to world coordinates
(3D). But I feel like it may be not applicable to my problems here, please
correct me if I got anything wrong.

The cortical surface reconstruction process will generate 2D surfaces from
the original 3D anatomical data. Here I wanted to projected my 3D
activation maps of VT cortex onto the 2D surfaces so that I will get a set
of 2D cortical maps from brain activation data (something like this figure:
http://www.pnas.org/content/103/26/10046/F4.large.jpg). I'm planning to
treat these activation data as 2D images and perform some analysis on them.
I wonder if it is possible to do certain projection and extract the
corresponding 2D coordinates for each voxel of interest. Or is it necessary
to manually cut the surface to create flattened patch first?

Best,
Xixi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.alioth.debian.org/pipermail/pkg-exppsy-pymvpa/attachments/20150225/a69d2e2f/attachment.html>