On the relevance of time in neural computation and learning
Theoretical Computer Science
Neural Computation
Effects of synaptic synchrony on the neuronal input-output relationship
Neural Computation
| Hi-index | 0.00 |
It is commonly assumed that temporal synchronization ofexcitatory synaptic inputs onto a single neuron increases itsfiring rate. We investigate here the role of synapticsynchronization for the leaky integrate-and-fire neuron as well asfor a biophysically and anatomically detailed compartmental modelof a cortical pyramidal cell. We find that if the number ofexcitatory inputs, N, is on the same order as the number offully synchronized inputs necessary to trigger a single actionpotential, Nt, synchronization always increasesthe firing rate (for both constant and Poisson-distributed input).However, for large values of N compared toNt, "overcrowding" occurs and temporalsynchronization is detrimental to firing frequency. This behavioris caused by the conflicting influence of the low-pass nature ofthe passive dendritic membrane on the one hand and the refractoryperiod on the other. If both temporal synchronization as well asthe fraction of synchronized inputs (Murthy and Fetz 1993) isvaried, synchronization is only advantageous if either N orthe average input frequency, fin, are smallenough.