Hello,
i just tried to compute MEG forward model but it seems i have some surprisingly large vector:
I first had a warning about 2 dipoles outside the brain, but used the force dipole inside the skull function.
WARNING: 2 dipole(s) outside the BEM layer "Brain".
Edit: I recomputed the BEM surfaces with a higher number of vertex and it seems to be working better than forcing the dipole inside the skull. Should we recommend that option instead when computing the head model?
Edouard
Hi @edelaire
Thanks for the feedback.
Using more vertices will not always fix the dipole outside the skull. In your case, I assume it is more likely that the skull's surface changed slightly, and the dipole was then moved inside.
One solution is to move the skull slightly (align manually on… cortex), inflate the innerskull surface by few mm, or generate new set of BEM surfaces, or even.
If the problem persists with the BEM, another possible solution is to use the simplified FEM on the BEM surface. You can generate the FEM mesh from the BEM surfaces, and then call Duneuro as in this example:
