Simulated EEG to Simulated MEG

Dear Pr. Francois,

I would like to ask you if iy possible to convert the simulated electracal field of each cell in the brain
to simulated MEG,

Because in our lab we have a model for EHG (for the contraction of uterine activity) which is electrical activity
and I would like to analysis the source localization for MMG data (also for the contraction of
uterine activity but it is magnetic).

So if there is an equation to convert the electrical to magnetic simulation so I can use it.

Best regards
Saeed

Dear Saeed,

In theory, if you have really good forward models, you should be able to do what you describe.

  1. Simulate EHG signals
  2. Compute a EHG forward model
  3. Invert this model and estimate the uterine sources from the simulated EHG
  4. Compute a MHG forward model
  5. Compute the MHG simulated signals by multiplying the source activity with the MHG forward model.

There are many things that are not available in Brainstorm:

  • the accurate geometry of the MHG magnetic sensors
  • the EHG and MHG forward models
    You would have to develop this on your own…

Cheers,
Francois

Dear Pr. Francois,

I hope you are fine,

[B]There are many things that are not available in Brainstorm:

Thanks you for your suggestions,

Best regards
Saeed

The page you mentioned describes the position of the MHG sensors, but not the geometry of these sensors.
You will need it to be able to compute a forward model for them (size and type of the coils).

To be able to model correctly the sensors in Brainstorm, you would need to add an entry to the file :

If you are interested in sharing your work, you could start thinking about distributing an example dataset and a tutorial, similar to these examples:
http://neuroimage.usc.edu/brainstorm/Tutorials/Decoding
http://neuroimage.usc.edu/brainstorm/Tutorials/PLS
http://neuroimage.usc.edu/brainstorm/Tutorials/EyetrackSynchro

Dear Pr. Francois,

Thank you,

About sharing my work, I will be glad to share my work to Brainstorm, my work is to make source localization and connectivity analysis for uterine sources
in the purpose of prediction the premature labor. It is interested project in the mean of huminity since Premature birth is 9.6% of all births worldwide making preterm birth one of the world’s largest public health problem which leads to high morbidity and mortality of newborns. But now I am in a searching process for the suitable tool to solve the source localization and then the source connectivity, for this I am here in Brainstorm.

The SARA system consists of 151 primary magnetic sensors that are spaced approximately 3 cm apart over an area of 850 cm2. The attached
PDF may help, when I sloved the forward problem to get the EHG signals (electrical signals), I noticed that openmeeg software need just the position of the electrodes,
so I put just the poistion of the electrode, is this correct? regarding your below mail it require the size also, but I don’t know where in Openmeeg we can identify the size of sensor. And the type of coil in MMG, may be openmeeg don’t request the type also, thank you for clearing this point.

Best regards
Saeed


Saeed Khalil Zahran
UMR 7338 : BioMécanique et BioIngénierie (BMBI) - Room: D220
Université de technologie de Compiègne
Tel : (+33) 06 63230580

For the forward modeling of the electric potentials (EHG), you just need one 3D position for each sensor (the center of the electrode).

For the magnetic sensors (MHG), you also need the orientation of the coil to know along which axis the field is measured, and in which direction.
If the sensors are gradiometers, you need to know the position of the two coils.

For a more accurate modeling, Brainstorm uses the geometry of the sensor to estimate the leadfield at different points in space around the center of the sensor:
http://neuroimage.usc.edu/brainstorm/Tutorials/ChannelFile?highlight=(integration)#Edit_the_channel_file
This is optional but recommended.

Thank you very much!, Now I remember, last year I try to solve the forward probem using Openmeeg, with those sensors and I noticed that we have to put the orientation of each electrode, so I orient them toward the uterus mesh, and I get the meg leadfield… I will search in internet to see if I can find the manual of those sensors, so we can know the exact orientation. Otherwise I don’t know what should be a solution, and if correct to put manually the approximately orientation of each sensors.

Best regards
Saeed

You should really use the real orientations. Magnetic sensors are not like electrodes, they are very sensitive to the orientation.
The manufacturer should be able to give you this information.

I tried to contact the people who developed Uterine Magnetomyography which is patent technology about the information about the sensors (size and orientation)
http://www.google.ch/patents/US20030199749

but they didn’t answered me,

I read just one info, that the sensors are oriented 45 degrees

Oriented 45 degrees? Do you know what this means?
Do you know at least what type of sensors it is? (magnetometers, axial gradiometers, planar magnetometers)

Goodmorning Dr. Francois,

I hope you are fine,

The sensors array covers a region of approximately 45 cm high and 33 cm wide, with a surface of 1300 cm2 inclined at[B] 45 degree.[/B]
I found the above sentence in this link: http://www.ese.wustl.edu/~nehorai/research/ra/back.html

And it seems [B]magnetometers[/B] since when I write in google “uterine mmg magnetometers” I get the below link:
http://www.ese.wustl.edu/~nehorai/research/ra/MMG3.html
The uterine magnetomyogram (MMG) measures the magnetic fields associated with … of our distributed processing scheme using an array of [B]magnetometers[/B].

Best regards
Saeed

FYI: To address all the recent questions we had related with simulations, I added some code in Brainstorm and wrote a new tutorial:
https://neuroimage.usc.edu/brainstorm/Tutorials/Simulations

Just committed a new process "Simulate recordings from dipoles":
https://neuroimage.usc.edu/brainstorm/Tutorials/Simulations#Single_dipoles

image
image