Optimal d_lambda for measuring propagation speed
Posted: Mon May 02, 2011 9:49 am
I am trying to measure the propagation speed in the myelinated axon model of McIntyre et al. (2002, http://senselab.med.yale.edu/ModelDb/sh ... model=3810).
I'm measuring the propagation speed by detecting the time the membrane potential passes -40mV using an APcount point process (code can be provided). I am using CVODE with cvode.atol = 0.001. I am looking at nodes away from the end of the axon to elimate end-effects.
I wanted to check that lambda was set to a sufficiently small value. Using the fixnseg.hoc script at http://www.neuron.yale.edu/neuron/stati ... ambda.html, I have changed d_lambda, and observed the propagation speed computed by dividing the distance between 8 nodes by the time taken for the AP to travel between them. Altering d_lambda, I get the following result:
I had expected that as I decreased d_lambda, the answer would start to converge, but it seems to be diverging. Is this behaviour to be expected? Admittedly, with d_lambda small, the number of compartments is rather large and impractical. However, it would be good to know if there is a reason known for this behaviour. Is there a lower limit on d_lambda?
I'm measuring the propagation speed by detecting the time the membrane potential passes -40mV using an APcount point process (code can be provided). I am using CVODE with cvode.atol = 0.001. I am looking at nodes away from the end of the axon to elimate end-effects.
I wanted to check that lambda was set to a sufficiently small value. Using the fixnseg.hoc script at http://www.neuron.yale.edu/neuron/stati ... ambda.html, I have changed d_lambda, and observed the propagation speed computed by dividing the distance between 8 nodes by the time taken for the AP to travel between them. Altering d_lambda, I get the following result:
Code: Select all
d_lambda V (m/s)
0.002 43.45
0.005 49.31
0.01 54.25
0.02 58.81
0.05 55.69
0.1 57.39
0.2 57.39