Resegment MRI after cleaning of recordings

Discussions
1

Hi Team,
I had imported BIDS dataset with default 15000 vertices. I segmented the MRI using CAT12 and filtered, removed artefacts. I have few queries.

  1. Can I re-segment the MRI using 1000 vertices? How would it affect the cleaned raw recordings and the co-registration?
  2. I understand that these vertices are dipoles within each 15000 voxels. So vertices == voxels right ?
  3. If I clean the raw recordings first and then can I do MRI Segmentation and MRI - MEG Co-registration ?
    Thanks,
    Rasmita
2

The number of vertices is for the cortical surface, thus this will not affect at all the raw recordings, these are in the sensor space. The low-resolution cortical surface (1000 vertices) will be co-registered with the MRI, when the sensors are co-registered with the MRI, then they will be also with the surface.

There is a bit to unfold here. MRI data is comprised of voxels (volume pixels), when MRI data is segmented, the cortical surface is extracted, and this one is defined as a mesh of vertices in the space. The number of vertices that you indicate in Brainstorm when importing an anatomy is the desired number of vertices for the cortical surfaces.

When estimating sources, if the source space with distributed methods it is needed to define the location of the dipoles, these location can be constrained to the cortical surface (surface source space) or to the whole brain (volume source space). In case of the Surface source space, then, one dipole is assumed for each vertex in the cortical surface mesh.

MRI segmentation will not have any impact in the MEG raw recordings, one is anatomy the other sensor space. While the MRI-MEG co-registration can be done later in the pipeline, when MEG data is imported it has already spatial information about the sensor location, it is expected to perform the co-registration with anatomy early in the pipeline, as it will be used when creating the head model.