The relationship between cross-correlation measures and underlying synaptic events

The relationship between an excitatory postsynaptic potential (EPSP) and the cross-correlation derived from the spike train evoked by a Poisson stimulus train was studied by computer simulation. The neuron model had a constant threshold and an absolute refractory period. Following a spike, the membrane potential was determined by ongoing activity, i.e., it was not 'reset'. Under all conditions studied, the rise time of the cross-correlation peak accurately reflected EPSP rise time. The agreement between the half-width of the cross-correlation peak and that of the underlying EPSP depended upon the EPSP amplitude, the rate of the driving Poisson input, and the time course of the EPSP. Over a broad range of experimentally relevant conditions, the half-width of the cross-correlation peak deviated from that of the underlying EPSP by less than 30% and bore little resemblance to that of the EPSP's time derivative.