My issue is finding the "distance between the simulated dipole and current density dipole.
If I am not wrong, I need to compare MAPPED data i.e KERNEL of the simulation and estimate. I did find KERNEL for the estimate in bst_memory, but I am having trouble on finding the simulated data KERNEL.
I underestand that the simulated source is computed differently in brainstorm, but there still should be a matrix representing how data points sits on the 3D brain (provides information with regards to location of the active sources).
Any help on where I can find the simulation kernel OR a way to compute SD?
Currently I only have access to KERNEL of the Estimate, Source ESTIMATE itself and SOURCE SIMULATION. I dont get how finding the difference in distance is observable with the source information
Because you need to compute distances between sources, I think you need to access the SourceLoc array of 3D coordinates of the cortical sources used in your model. It should be stored in the head model file. You can then refer to the source Estimate and Simulation arrays for the source amplitudes you need to compute SD.
The kernel is an array applied to the sensor recordings to compute the source Estimate array.
Hope this helps.
Thank you Dr.Sylvain for the reply.
So from what I understood, there is no specific matrix in Brainstorm that have information of Location and Amplitude together?
I thought what the matrix that has been used to "illustrate" brain surface cortex in 3D image is the matrix that I can use to compute distance between sources.
Do you mind explaining how does
I am so sorry but I am still confuse on finding location of the sources ...
Thank you so much. I was able to find the matrix in both Verticies and GridLoc of surface , both are same and 15,000 x 3 matrix. However, If I am not wrong, they are representing the 3D brain without consideration for the sources.
How I can now map the Estimated source , or Simulated Source on this 3D surface?
i.e for simulation all points on 3D matrix must be zero except few locations where sources are.
I think I figured it out … All I need to do is to observe the simulation result, and just compare the columns with the same 15,000 vertices from the GridLoc.
i.e, any point in source from 15,000 that is zero, will lead to same point in GridLoc to be zero.
The Vertices/GridLoc matrix contain the 3D positions of the vertices of cortex surface, which are also the 3D positions of the dipoles in your distributed source model (in the case of the surface-based source space). The positions of the dipoles are fixed, and the activity at each of these points is estimated.
