Challenges in EEG Source Localization

[sarach]

Hi

I'm currently focused on studying the visual cortex. I've gathered all the necessary MRI data in the freesurfer folder and then Brainstorm. I've successfully created a head model using BEM and openMEEG, even though I didn't have access to matching EEG data. Instead, I used a default EEG cap to build the head model.
Based on brainstorm tutorial, I used this code: Gain_constrained = bst_gain_orient(HeadModel.Gain, HeadModel.GridOrient);
The dimension of the output matrix is three times smaller : [Nsensors x Nvertices]
Next, I calculated the sum of the 'Gain_constrained' values along the second axis, resulting in a 128x1 array. When I visualized this data, I generated a topomap. However, the topomap appears to be inaccurate. It's important to note that my current focus is on a specific region of interest (ROI), and in this context, I'm particularly puzzled by the behavior of the 'Gain_constrained' values.
Can you please help me identify what might be going wrong in my work?

[Sylvain]

I am not sure what you are trying to accomplish here, sorry. Please note that the gain matrix is data independent: it is based on the geometry of the head, locations and orientations of the brain sources and locations of the EEG scalp electrodes. Whether you are deriving the gain matrix /head model in the context of a vision study does not make a difference.
Instead, you want to visualize the source maps, which are obtained by combining the actual EEG recordings with the head model. Please refer to the Brainstorm tutorials to obtain those maps, and let us know if you obtain the expected basic visual responses after the onset of the stimulus presentation, for example.

[tmedani]

I also advise as @Sylvain recommendations.

regarding your issue here:

@sarach

Gain_constrained = bst_gain_orient(HeadModel.Gain, HeadModel.GridOrient);
The dimension of the output matrix is three times smaller : [Nsensors x Nvertices]

The bst_gain_orient constraints the orientation of your dipole to the values within HeadModel.GridOrient,
So this explains why your Gain_constrained is 3 times smaller.