www.neuron.yale.edu

(In response to
Extracellular stimulation and recording
viewtopic.php?t=168)

Dear colleagues,
I'm working on a direct current model of transcutaneous electrical stimulation of the spinal cord. The volume conductor model was thought to be purely resistive, with multiple layers of tissues (skin, muscles, bones, etc.). COMSOL was used to simulate the electric field induced in the spinal cord. In this way, I hope to simulate how the field affects the excitability of motoneurons (particularly somata and dendrites). My question is, how can I apply DC stimulation using the "extracellular" and "xtra" mechanisms? Will the distribution of the electrical potential calculated with the FEM over the conductive volume (gray matter) in this case be the input for the model's extracellular electrical potential? Or should I rely on the electric field? Should I calculate rx using V (Calculated Electric Potential) / I (input current), which is defined as "transfer resistance between the stimulus electrode and the local node"?

darkleo wrote: ↑Tue Dec 12, 2023 9:04 amdirect current model of transcutaneous electrical stimulation of the spinal cord. The volume conductor model was thought to be purely resistive, with multiple layers of tissues (skin, muscles, bones, etc.). COMSOL was used to simulate the electric field induced in the spinal cord.
. . . how can I apply DC stimulation using the "extracellular" and "xtra" mechanisms? Will the distribution of the electrical potential calculated with the FEM over the conductive volume (gray matter) in this case be the input for the model's extracellular electrical potential?

Yes.

Should I calculate rx using V (Calculated Electric Potential) / I (input current), which is defined as "transfer resistance between the stimulus electrode and the local node"?

You could, if that's convenient. In this case, you would have told COMSOL the shape and location of the stimulating electrode, and the time course of the applied current, and used COMSOL to calculate the time course of extracellular potential in the conductive medium, then interpolated the extracellular potentials at the locations that correspond to the centers of your model's segments). Then, the transfer resistance between any given location (x,y,z) and the stimulating electrode would be rx(x, y, z) = v(x, y, z, t)/i(t). You want v and i to be in units such that rx will be in megohms, e.g. mV and nA, or V and uA. If the tissue is nondispersive (i.e. purely resistive), then the values of rx can be calculated using a very simple, brief waveform--for example, a rectangular pulse starting at t = 1 mS that lasts 1 mS.

You might find it useful to read this post viewtopic.php?p=20130#p20130.

You might also want to look at this article
Reilly JP. Survey of numerical electrostimulation models. Phys Med Biol. 2016 Jun 21;61(12):4346-63. doi: 10.1088/0031-9155/61/12/4346. Epub 2016 May 25. PMID: 27223870.
Let me know if you have difficulty getting a copy of this article.
Source code for the NEURON model is available from https://modeldb.science/239006
This illustrates use of xstim (similar to xtra but just a bit easier to use because it omits the stuff that calculates local field potential generated by neuronal activity).
Also shows how to use fzap.mod and fsquare.mod as function generators to drive extracellular potential.

--Ted