Projection kernel on default anatomy

1

Dear BST users,

I’ve a quick question:

As suggested in the forum I’ve extracted an MNE inversion kernel for individual anatomies for computing single-trial time-series at source level on the fly.
Now I’ve run this analysis for each subject and I need to project each trial sources time series on default anatomy.
Is there a way to “project the kernel” instead of projecting single trial time courses?
This will save a lot of time in my analysis.

Thanks for your help
andrea

2

No, we removed this option because it was too confusing and too complicated to handle.
The recommend solution is to compute a subject-level average for each subject, project these averages on the template, and then run your group statistics:
http://neuroimage.usc.edu/brainstorm/Tutorials/Workflows#Constrained_cortical_sources
http://neuroimage.usc.edu/brainstorm/Tutorials/VisualSingle
http://neuroimage.usc.edu/brainstorm/Tutorials/VisualGroup

4

Thanks François for your reply.

Actually I’m running a classification analysis at source level across subjects:
for this purpose I can’t work with the subject-level averaged trials but I need to project the entire time series on the default anatomy.
Any suggestion about bst_function that implements a “projection kernel” ?

andrea

5

Hello!
I am trying to project indidual imaging kernels on default anatomy.
The result is finally found in the Group analysis folder. A year ago I could extract the projected ImagingKernel from the result and work with it myself. But now the result has an empty Imaging kernel matrix

image
image1280×800 216 KB

6

Hello,

Indeed, the function to compute and apply the interpolation matrix between the two surfaces changed and now the ImagingKernel is never kept.
https://github.com/brainstorm-tools/brainstorm3/blob/master/toolbox/math/bst_project_sources.m#L226
https://github.com/brainstorm-tools/brainstorm3/blob/master/toolbox/math/bst_project_sources.m#L324

However, this changed occured more than 3 years ago. The first version of this function we uploaded on github in 2016 was already working in this way:
https://github.com/brainstorm-tools/brainstorm3/commit/960f8af6c732b60fb9def5115930670d0d1a1809#diff-46507fb910c215476f83065bc2f6960aR224
https://github.com/brainstorm-tools/brainstorm3/commit/960f8af6c732b60fb9def5115930670d0d1a1809#diff-46507fb910c215476f83065bc2f6960aR322