Hi,
Trying to have much more convenient deep sources (and not only superficial sources du to the BEM surface), I chose the MRI volume option when I compute my head model. However, when I try to run wMEM using that head model type computing from OpenMEEG, It doesn't work (see the error box I joined). Could you explain to me how to use correctly OpenMEEG with MRI volume option in order to apply the wMEM method (or any non linear methods) ?
Best regards,
I'm not sure if the MEM code is compatible with volume source models.
Let's ask the developers of the method:
@youneszer @klack @RChowdhury @tanguy @Lina @cgrova @aliobaibk?
Hello,
I confirm that the MEM code is not yet compatible with volume head models.
Thanks for your answer. How to obtain volumique sources using MEM methods ?
Best regards,
I added an extra test to return a clearer error message when trying to run the MEM on volume source models (it was already disabled from the GUI):
https://github.com/brainstorm-tools/brainstorm3/commit/def370851ea2fdfad3a99fa18a82e2e9f19bf14d
MEM and wMEM are based on a parcelization of the whole cortical surface, based on geodesic metric along the cortical surface. Therefore, an extension of MEM in a volume space is not straightforward, we are planning to implement a version of it but this will take some time, so not available for the moment. What can be done is extrapolating within the gray matter the recontruction done on the surface, this is the way we are usually providing results to our clinician colleagues. Hope this helps
Thanks for your answer.
I have the same question as marieLR. I would like to know if there are any developments in the implementation on the MRI volume. Alternatively, I would like to ask for information on the extraction procedure you mentioned. Could you please give me more details?
Thanks for your message, actually recruiting new fellows is a bit difficult those days and we are in progress, when my next postdoc will be able to arrive in Montreal, I think we will start investigating MEM in volume, but not for now, so it will require some more time. Regarding the "extraction" it is just a representation prepared by Francois et al, when you have your source loc results just chose the option "Display on MRI (MRI viewer) and MEM results will appear on the grey matter in 3D viewer, hope this helps, we will keep you aware of our next developments, Christophe
Hi Christophe, thank you very much for your reply. Yes, please I am very interested in the developments of MEM approaches in the MRI volume. In these weeks I have tested almost all the methods of source localization and the MEM, for my purposes, is the best because it is very focal and free of spurious activations as happens in the other methods. Unfortunately, the study I am conducting requires volumetric source localization and therefore I will proceed with the methods already available. Best, Margherita
