How to find the values for the Motor Cortex after the inverse solution?

Discussions
1

Hello, I’m fairly new to Brainstorm and currently need a little bit understanding and figuring a way for multiple problems.

  1. I have the GridLoc points from the .mat file but I want to be able to separate the points located in the motor cortex, I have the sMRI for the data or the MNI area of the cortex. What would be the best approach?

  2. I want to find the ImageGridAmp to compute a matrix from [Time by Power] and sum and square the results to get the amount of power in the motor cortex, but after computing the source with an unconstrained dSPM head model it shows up as 0x0 in the .mat file. Also how would i presume to read it?

  3. Also a further explanation on the results of computing the sources after the head model for mri volume and surface head modeling, such as the differences?

  4. For data that used the default B-152 template as the anatomy is Loreta or dSPM better used?

Thank You Very Much

2

I have the GridLoc points from the .mat file but I want to be able to separate the points located in the motor cortex, I have the sMRI for the data or the MNI area of the cortex. What would be the best approach?

Create a scout or use an existing one from one of the available atlases.
You'll find the indices of the vertices in the ROI in the surface file:
http://neuroimage.usc.edu/brainstorm/Tutorials/Scouts#On_the_hard_drive

I want to find the ImageGridAmp to compute a matrix from [Time by Power] and sum and square the results to get the amount of power in the motor cortex, but after computing the source with an unconstrained dSPM head model it shows up as 0x0 in the .mat file. Also how would i presume to read it?

To read the field ImageGridAmp? In a source file, you have either ImagingKernel or ImageGridAmp defined:
http://neuroimage.usc.edu/brainstorm/Tutorials/SourceEstimation#On_the_hard_drive

In a time-frequency file (you are talking about power: have you computed the power in different frequency bands?), the data is stored in the field TF:
http://neuroimage.usc.edu/brainstorm/Tutorials/TimeFrequency#On_the_hard_drive

Also a further explanation on the results of computing the sources after the head model for mri volume and surface head modeling, such as the differences?
For data that used the default B-152 template as the anatomy is Loreta or dSPM better used?

I cannot provide much more information than what is on these pages:
http://neuroimage.usc.edu/brainstorm/Tutorials/SourceEstimation
http://neuroimage.usc.edu/brainstorm/Tutorials/TutVolSource
You can also look for additional references in the literature to guide you in your analysis:
http://neuroimage.usc.edu/brainstorm/Pub

Please get back to us if you have more specific questions.

Cheers,
Francois

3

Hello Francois, thanks for the help but another problem I have is that:

  1. Iam wondering since we dont have ImageGridAmp as it shows empty then the result of the inverse solution is ImaginingKernel, so would that be the k value? Would ImageGridAmp be the J value underneath in .mat for the results?

2)Is there a way to rather get ImageGridAmp rather than ImagingKernel to get a time series value, as of a function or other manner ?

4

ImageGridAmp = ImagingKernel * Recordings
(Recordings = F field in the imported epochs and averages)

In the advanced options of the minimum norm, you have an option “Full results”, which does what you want.
It is available only if you try to estimate the sources for one file (otherwise, it would be necessary a shared inverse model, saved in the ImagingKernel matrix)