Inverse first passage time method in the analysis of neuronal interspike intervals of neurons characterized by time varying dynamics

We propose a new method to analyse time series recorded by single neuronal units with the aim to recognize the times when the neuronal interspike interval corresponds to different time evolutions. The effect of different dynamics is artificially concentrated in the boundary shape in a stochastic leaky integrate and fire model to allow the use of the inverse first passage time method. A suitable time window fragmentation on the observed data and the repeated use of the inverse first passage algorithm allows to recognize the existence of an evolution in the dynamics. The comparison of the boundary shapes in the different time windows detect this evolution. A simulation example of the method and its biological implications are discussed .