On the description of neuronal output properties using spike train data.

Neuronal output properties for input stimuli that evoke a deterministic response can be efficiently described by the interspike-interval function (Awiszus 1988a). It is shown in this paper that there are stimuli for which both the Hodgkin-Huxley (HH-) model of an action potential encoding membrane (Hodgkin and Huxley 1952) and a muscle spindle primary afferent generate responses which violate the conditions for a deterministic one. Instead of being stochastic these responses follow systematic rules, namely those for a semi-deterministic response, a class of neuronal responses established in this paper that includes the deterministic one. Instead of being stochastic these output properties are best described by the interspike-interval curve. A phase plane analysis of the internal properties of the HH-model underlying such responses shows that it is reasonable to assume that responses of an HH-model and consequently, all neurons for which an HH-model is a valid description of the action potential encoding process, always fall into the class of semi-deterministic responses, regardless of the input current density time course as long as it is large enough to maintain spike activity. Consequences of this assumption for the analysis of neuronal output properties are discussed with respect to output measures and efficient input stimuli.