Simplified models of neocortical pyramidal cells preserving somatodendritic
Neurocomputing 38: 167-173, 2001.
Simplified models are needed for performing large-scale network simulations
involving thousands of cells. Ideally, these models should be as simple as
possible, but still capture important electrotonic properties, such as voltage
attenuation. Here, we propose a method to design simplified models with
correct voltage attenuation, based on camera-lucida reconstructions of
neurons. The simplified model geometry is fit to the detailed model such that
it preserves: (i) total membrane area, (ii) input resistance, (iii) time
constant and (iv) voltage attenuation for current injection in the soma.
Using the three dimensional reconstruction of a layer VI pyramidal cell, we
show that this procedure leads to an efficient simplified model which
preserves voltage attenuation for somatic current injection as well as for
distributed synaptic inputs in dendrites. Attenuation was also correctly
captured in the presence of synaptic background activity. These simplified
models should be useful for performing network simulations of neurons with
electrotonic properties consistent with detailed morphologies.
return to main page