vext in extracellular mechanism

Anything that doesn't fit elsewhere.
Post Reply
cobaltplusone

vext in extracellular mechanism

Post by cobaltplusone »

Per the documentation on this page:

http://www.neuron.yale.edu/neuron/stati ... ml#index-5

vext[0] refers to the extracellular potential just outside the membrane. I cannot, however, figure out what the purpose of vext[1] is. There is a note that states, "The only reason the standard distribution is built with nlayer=2 is so that when only a single layer is needed (the usual case), then e_extracellular is consistent with the previous documentation with the old default nlayer=1." What does this mean? And what is the difference between xc, xc[1], and between xg, xg[1]?

Thanks in advance :)
ted
Site Admin
Posts: 6289
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: vext in extracellular mechanism

Post by ted »

cobaltplusone wrote:There is a note that states . . . What does this mean?
It means that legacy user code that employed extracellular and followed documentation written many years ago will work properly. Doesn't mean a hill of beans (i.e. has no implications) for new user-written code.
what is the difference between xc, xc[1], and between xg, xg[1]?
xc is a hoc nickname for xc[0], i.e. the capacitance in the extracellular layer that is immediately adjacent to the cell membrane. xc[1] is the capacitance in the outer extracellular layer. Similar comments apply to xg and xg[1].
luciana
Posts: 17
Joined: Sun Mar 09, 2014 2:29 pm

Re: vext in extracellular mechanism

Post by luciana »

Hi everyone,

I have another question related to Vext in extracellular mechanism.
I want to adjust the amplitude of the extracellular action potential simulated to our experimental data. I thought for that purpose I should change xg parameter (conductivity in extracellular layer) and the waveform of extracellular action potential wouldn't change, but it does. Why is that?
Is not true that:
Vext = I_tranmembrane / extracellular conductivity (aprox.) ?

Thank you all.
Luciana.-
ted
Site Admin
Posts: 6289
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: vext in extracellular mechanism

Post by ted »

This is really a separate question and belongs in its own discussion thread. After you have a chance to read this reply, I will have to come back to this thread and split it in two.
I want to adjust the amplitude of the extracellular action potential simulated to our experimental data. I thought for that purpose I should change xg parameter (conductivity in extracellular layer)
Nice thought, but no, that is not the way to proceed. It would be a meaningless parameter change, something that would produce a nonsense result.

Why? extracellular is not a representation of the electrical properties of the extracellular milieu. It is primarily a tool for providing users direct access to the external surface of neurites for the purpose of applying extracellular stimuli. It can also be used to represent a myelin sheath when it is necessary to also represent the properties of the axolemma. But it is definitely NOT a substitute for a Poisson solver (a tool that solves the Poisson equation, a PDE that specifies how potential in a volume conductor depends on the complex admittance of the volume conductor and current sinks and sources that are distributed in that volume conductor).

That said, if the extracellular medium is purely resistive, then the potential at any point p can be calculated as the instantaneous sum

Code: Select all

SUMMA rx_j * i_j
     j
where i_j is the current generated by the jth current source and rx_j is the transfer resistance between p and the location of the jth current source. This is the approach used by LFPy (Linden et al. 2013, Front. Neuroinformatics 7, PMID 24474916) and also by the code I developed that is described in the Hot tips area of this Forum under the topic Extracellular stimulation and recording.
But neither LFPy nor my own code is a Poisson solver, and neither of them can be used to examine what would happen to a model cell if the resistive properties of the extracellular medium were changed. Neither of these software packages takes into account the effect of extracellular impedance on membrane potential, or enables construction of models that involve ephaptic coupling between different neurites.
canavier
Posts: 4
Joined: Mon Apr 24, 2017 10:15 am

Re: vext in extracellular mechanism

Post by canavier »

What we are trying to figure out is why Vext (and the LFP Luciana measures in LFPy) is so tiny when the actual measured extracellular potentials are on the order of a millivolt. We don't really care about the resistance of the extracellular medium, we are trying to figure out why the voltage measured by the extracellular electrode is so much greater than the drop between the extracellular potential next to the membrane. I think our fundamental error it is that looking at a single point outside the membrane and the current flow relative to a single segment doesn't take into account all the current flow induced by the wild inhomogeneity of the membrane potential in a spatially distributed neuron during an action potential. However, we need to compare the computed LFP with the experimentally recorded ones. Assuming a purely resistive medium, LFPy also gives small numbers, but Luciana please correct me if I remember wrong. I thought LFPy summed all the currents as you describe, so it should give a decent sized LFP. Is it possible we underestimate the axial (and therefore the transmembrane) currents because our Ra and other passive properties are off? Or maybe we just need a faster rate of rise on the AP?
ted
Site Admin
Posts: 6289
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: vext in extracellular mechanism

Post by ted »

canavier wrote:What we are trying to figure out is why Vext (and the LFP Luciana measures in LFPy) is so tiny when the actual measured extracellular potentials are on the order of a millivolt.
I can't help with LFPy, but it shouldn't be hard to do a sanity check with my own code for simulating extracellular potentials. Is your model a single model cell, and how is the cell being stimulated?
I think our fundamental error it is that looking at a single point outside the membrane and the current flow relative to a single segment doesn't take into account all the current flow induced by the wild inhomogeneity of the membrane potential in a spatially distributed neuron during an action potential.
I thought LFPy took care of all that stuff.

To run a "sanity check simulation" for you, I'd need the hoc, ses, and mod files that you use to set up the properties of your model cell (not in the Forum, but by email to me as a zip file), a short description of how you're simulating it--e.g. with an IClamp, or by one or more synaptic inputs etc.--and where in extracellular space you are "recording" the extracellular potential.
canavier
Posts: 4
Joined: Mon Apr 24, 2017 10:15 am

Re: vext in extracellular mechanism

Post by canavier »

Dopamine neurons are pacemakers, so they do not need to be stimulated. Besides, the extracellular recordings
we are trying to mimic are in vivo. I do appreciate the offer of help, but I'll let Luciana follow up since she is doing the simulations.
luciana
Posts: 17
Joined: Sun Mar 09, 2014 2:29 pm

Re: vext in extracellular mechanism

Post by luciana »

Thank you very much Ted. I will send you the hoc,mod and ses files to your e-mail.
Post Reply