Dendritic transformations on random synaptic inputs as measured from a neuron's spike train--modeling and simulation.

Extracellular spike trains recorded from central nervous system neurons reflect the random activations from a multitude of presynaptic cells making contacts mainly on the extensive dendritic trees. The dendritic potential variations are propagated towards the trigger zone where action potentials are generated. In this paper, two dendritic propagation modes are modeled: passive and quasi-active. Synaptic bombardments are modeled as being applied apically, somatically, or distributed over the dendritic tree. The resulting simulated neuronal spike trains are analyzed by point process techniques. Dendritic inputs resulted in a tendency for random bursting, interspike interval histograms with a long tail and coefficients of variation larger than one. The autocorrelation histograms reflected dynamics of the dendritic tree and they were able to discriminate between a passive or a quasi-active propagation mode and between dendritic and somatic synaptic inputs.