Applied current as a function of time

The basics of how to develop, test, and use models.
Post Reply
physicist

Applied current as a function of time

Post by physicist »

Hello,
I've found papers that use the NEURON for this, but I haven't succeeded to make NEURON do that :) So, I need to apply stimulus current to the axon, in such way that it is not a constant direct current of 1ms and so on... I need it to be alternating sinusoidal current, or periodic biphasic current of high-frequency.
I'm an undergrad student and I want to play a bit with nerve block simulations, cause I read some impressive papers on it :)
ted
Site Admin
Posts: 5810
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: Applied current as a function of time

Post by ted »

Ever done any programming?
physicist

Re: Applied current as a function of time

Post by physicist »

I have. So it has to be done through coding (hoc or wte)?
Thanks a lot :)
ted
Site Admin
Posts: 5810
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: Applied current as a function of time

Post by ted »

Could be done by programming, could be done with the GUI.

The simplest first step is to examine the effect of sinusoidal injected current on a single compartment model that has the Hodkgin-Huxley mechanism. In broad outline, the steps are:
1. Set up a single compartment model driven by an IClamp and demonstrate that it makes a spike.
2. Attach a point process that delivers a sinusoidal current. You could use Izap for this, with f0 = f1. You'll find a link to the NMODL source code for Izap at http://www.neuron.yale.edu/phpBB/viewto ... f=8&t=2404

You can do both 1 and 2 with the GUI. For 1, use NEURON Main Menu / Build / single compartment
and click on the single compartment's hh button. Also bring up a RunControl panel, a Voltage axis graph, and a Point Process manager configured as an IClamp with del = 1 ms, dur = 0.1 ms, amp = 0.1 nA. Adjust IClamp.amp until you are just at spike threshold. With a binary search strategy, you should be able to get 3 place accuracy in just a handful of trials--10 or so.

For 2, just bring up another point process and configure it to be an Izap. You'll want to change tstop and IClamp.del to values much larger than 1 ms (need to make sure that there has been sufficient time for your applied sine wave to go through at least 3 or 4 cycles). Also need to make sure that your test stimulus pulse--the one you're going to use to demonstrate whether the model is more or less excitable--is applied at a time when Izap's current is 0. You'll want to test for excitability on both the downstroke and upstroke phases of the injected current, so you'll want to bring up a current axis graph that displays Izap.i
Post Reply