Different solutions computing sources by hand or via the Process tab

1

Hello Brainstorm team,

We are using the "Compute Sources [2018]" function with Minimum norm imaging, sLORETA, and Unconstrained settings on one condition that contains 30 averaged trials.

When we display the cortical activation on the cortex, the solutions look different depending on whether we computed the sources by:

  1. right clicking the averaged trials folder or;
  2. dragging the averaged trials folder to the Process tab > Run>Sources>Compute sources [2018] with same settings.

(see attached image)
The first image depicts the kernel solution projected on the cortex when Option 1 is done.

#1RightClick
#1RightClick726×595 107 KB

The second image depicts the solution when the computation is done using the Process tab.

#2ProcessTab
#2ProcessTab720×601 140 KB

Please advise.

Thank you very much.

2

Both versions are computed using the same function.
If they give different results, either they used different head models or different options.

If you can't find the difference: put a breakpoint at line 145 of file bst_inverse_linear_2018.m, start the computation, and explore the contents of the variable.
If you have never used the Matlab debugger, watch a short introduction video about it, it is simple and very useful tool.