Hello, I am planning an experiment and am looking to use Brainstorm to process sEEG data in a rodent model for source localization. Regarding the need for anatomy data for the subjects, I have a CT scanner more readily available than an MRI and so I am curious as to whether both the pre- and post- implantation scans can be done using CT imaging or if the initial scan must be MRI.
Also, I just want to confirm that the a pre-implantation scan is necessary to generate the required volumes/surfaces and source localization can't be done with a post scan alone.
I am curious as to whether both the pre- and post- implantation scans can be done using CT imaging or if the initial scan must be MRI.
There is no restriction for the type of volumes you import: CT or MRI. You can import 2 CT volumes if you need to. The only necessity is to set 3 fiducial points (NAS, LPA, RPA - even arbitrarily) in the first imported volume, as these define the "Subject Coordinate System", used for the location of the electrodes. https://neuroimage.usc.edu/brainstorm/CoordinateSystems
I just want to confirm that the a pre-implantation scan is necessary to generate the required volumes/surfaces and source localization can't be done with a post scan alone.
You need a volume that some tools can process in order to produce the different surfaces you need for the source estimation (cortex, inner and outer skull).
I don't know any program that can perform an automated segmentation of a rodent MRI. Do you?
Thank you for getting back to me; this information is helpful. At present, I don't know of any automated rodent segmentation so I'm of the impression that I will have to do it manually through 3DSlicer and the like or create a system myself.
I see in the tutorial data set that there are quite a few individual volumes/surface parcellations in this anatomy object. Does this pipeline work with just 1 set of surfaces for the cortex, inner, and outer skull or are the individual subdivided volumes also needed?
Realistically, I just need a location in 3D space so I could also just model a 3D mesh of the relevant volumes by hand and use those if the software can handle that type of input.
Only the surfaces: scalp, outer skull, inner skull, cortex.
I don't know the feasibility of using these tools on rodents, I don't know if this has been done before...
I've seen some publications that use the FEM for TES/TMS for monkeys, but not for rodent.
But for the EEG/iEEG modeling part , as long as the input data is in the correct format, you can perform the FEM computation. You need to have the mesh of your model as one piece (either 1 tissue or multiple tissues), and all the electrodes should be on the mesh (EEG case) or inside the mesh (iEEG case).
You need also to define your source space and the conductivity values for your tissue(s).
If you have your set of surfaces for the cortex, you can generate the mesh with the iso2mesh as explained on the tutorials.
Please let us know how it works.
I know that Brainsuite can be used with some non-humain brain, but need to confirm this with @Anand_Joshi
Once you have the segmentation, it's straight forward to get the tessellation.