Numbers in leadfield matrix

[haniehsotudeh]

Hi Francois
i have created headmodeliing based on the tutorials in Brainstrom, but the numbers in obtained leadfield matrix have been unusual for me, because they greater than one, i think that they should be smaller than one. for example if we have some active sources in the brain and if we want to calculate their effects on EEG electrodes, with mentioned numbers in leadfield matrix, we access the great number for source effects on electrode, while the effects of inner source should be decrease when passes the three layers of the head... can you help me? is these number true... the obtained leadfield matrix related to the EEG-epilepsy is indicated below...

[Francois]

The values you can find in the Gain matrix of the headmodel file are not in the same units as the measured scalp EEG values. This is explained here:

EEG units

HI, I am curious about the units used for forward head modeling. Are they supposed to be microvolts on the scalp surface? I tried three different cortical source models as follows: dipoles min forw soln max forw soln from 2,503 -1,912 2,016 Freesurfer (lh & rh.pial) 10,003 -2,041 2,016 FreeSurfer (" ") 64,000 -0.0028 0.0026 Human Connectome Projec…

[haniehsotudeh]

thanks, but i want to ask question that if i want to create simulated sources in the brain and then calculate the simulated EEG data on the surface, for solving inverse problem, are these units in gain matrice cause a problem or not?
can it used for solving the inverse problem or we should changed its units? i mean that Can it be used in the same way?

[Francois]

If you want to run simulations, you can refer to the tutorial Simulation:
https://neuroimage.usc.edu/brainstorm/Tutorials/Simulations

[tmedani]

@haniehsotudeh

i have created headmodeliing based on the tutorials in Brainstrom, but the numbers in obtained leadfield matrix have been unusual for me, because they greater than one, i think that they should be smaller than one.

The values on the leadfield matrix are the values of the simulated voltage or the forward field from a given source at a given channel with regard to a reference channel.

in your matrix you have 41x45006, so your model has 45006/3 = 15002 dipoles and 41 channels. when computing the leadfield matrix each dipole is simulated in the 3 directions (x, y and z).
you can check the structure on this page:
https://neuroimage.usc.edu/brainstorm/Tutorials/HeadModel?highlight=(gain)|(matrix)#On_the_hard_drive

If you want to simulate one dipole, you can refer to the simulation tutorial.
If you want to check how the leadfield is computed in basic sphere model you can refer to this function: brainstorm3/toolbox/forward/bst_eeg_sph.m at master · brainstorm-tools/brainstorm3 · GitHub

[Francois]

@tmedani

in your matrix you have 41x45006, so your model has 45006/3 = 15002 dipoles and 41 channels

To define the terminology more precisely: there are actually 45000 dipoles in your model.
15000 different locations in space (the vertices of the cortex surface), with three orthogonal directions (x,y,z) are each point, for a total of 45000 dipoles.