|
The spiking patterns of neurons can be often highly variable. In many cases, the origin of this neural noise is not known and might be difficult to access directly. Here, we explore the possibility to distinguish between two different kinds of intrinsic noise solely from the interspike interval (ISI) statistics of a neuron. Specifically, we consider two adaptive neuron models in which fluctuations (channel noise) are either associated with fast ionic currents or with slow adaptation currents that are observed in many neurons. We show by means of analytical techniques and extensive numerical simulations that the shape of the ISI histograms and the ISI correlations are markedly different in both cases: For fast current fluctuations and deterministic adaptation, the ISI density is well approximated by an inverse Gaussian and the ISI correlations are negative. In contrast, for stochastic adaptation currents, the density is more peaked and has a heavier tail than an inverse Gaussian density and the serial correlations are positive. For the theoretical analysis we use an analytically tractable integrate-and-fire model. The results are qualitatively confirmed by simulations of a biophysically more realistic Hodgkin-Huxley type model. Our results could be used to infer the dominant source of noise in neurons from their ISI statistics. |
|