Fig. 13.

Percent discrimination of a sinusoidally amplitude-modulated pulse train. Three discrimination measures are were used to discriminate unmodulated pulse train from a modulated pulse train on the basis of simulated spike trains (see text for details). Both stimuli were one second long trains of biphasic pulses, 40 μs per phase and current level was varied for each carrier pulse rate so that the unmodulated pulse train always evoked ∼50 spikes/s, on average. Modulated pulse trains had 75-Hz modulation frequency and 1% modulation depth. Percent correct values were computed from 1,000 repeated simulations of the model, error bars represent SE in the mean of 10 repeated estimates of percent correct. Chance level is 50% correct.