Encoding Chaos in Neural Spike Trains

A recent theoretical study showed that with a chaotically driven integrate-and-fire neutron model, there can be a one-to-one correspondence between the system states of the chaotic input and the interspike intervals (ISIs) from the model. In this investigation, an attempt is made to extend the previous work to in vitro studies on rat cutaneous afferents. In particular, the hypothesis that the deterministic structure of a chaotic input signal can be preserved when the signal is converted into a spike train by a sensory neuron is tested. Ten neurons are included in the investigation. Determinism of the ISI series is assessed using a nonlinear prediction algorithm.