Non-stationary filtered shot noise processes and applications to neuronal membranes

TL;DR

This paper develops a comprehensive analytical framework for non-stationary filtered shot noise processes, particularly applied to neuronal membrane potential modeling, enabling precise analysis and estimation of neural response statistics under variable input rates.

Contribution

It introduces exact and approximate formulas for joint cumulants of non-stationary filtered shot noise, and applies these to neuronal models to estimate membrane potential dynamics and pre-synaptic activity.

Findings

Derived exact expressions for joint cumulants of non-stationary shot noise.

Provided accurate approximations for membrane potential distribution over time.

Developed a simple method to estimate pre-synaptic rate from membrane potential traces.

Abstract

Filtered shot noise processes have proven to be very effective in modelling the evolution of systems exposed to stochastic shot noise sources, and have been applied to a wide variety of fields ranging from electronics through biology. In particular, they can model the membrane potential Vm of neurons driven by stochastic input, where these filtered processes are able to capture the non-stationary characteristics of Vm fluctuations in response to pre-synaptic input with variable rate. In this paper, we apply the general framework of Poisson Point Processes transformations to analyse these systems in the general case of variable input rate. We obtain exact analytic expressions, and very accurate approximations, for the joint cumulants of filtered shot noise processes with multiplicative noise. These general results are then applied to a model of neuronal membranes subject to conductance shot noise with continuously variable rate of pre-synaptic spikes. We propose very effective approximations for the time evolution of Vm distribution and simple method to estimate the pre-synaptic rate from a small number of Vm traces. This work opens the perspective of obtaining analytic access to important statistical properties of conductance-based neuronal models such as the the first passage time.