Successive spike times predicted by a stochastic neuronal model with a variable input signal

Area of Science:

• Computational Neuroscience
• Mathematical Biology
• Stochastic Processes

Background:

• Neuron membrane potential dynamics are complex and influenced by time-varying signals.
• Accurate modeling of neuronal spiking is crucial for understanding neural computation.
• Stochastic processes offer a powerful framework for analyzing neuronal excitability.

Purpose of the Study:

• To develop and compare stochastic models for neuronal membrane voltage evolution.
• To accurately predict the timing of successive neuronal spikes.
• To provide a generalizable framework for modeling spike trains under time-varying input.

Main Methods:

• Modeling membrane voltage with an inhomogeneous Ornstein-Uhlenbeck process.
• Utilizing a Gauss-Markov process, dependent on the first passage time of the first process.
• Analyzing the probability density function of the maximum first passage time for spike time approximation.
• Comparing simulation results with numerical and asymptotic approximations.

Main Results:

• Demonstrated that the second Gauss-Markov process is also a diffusion type.
• Developed an approximation for the second spike time distribution using probability density functions.
• Provided a general equation for modeling successive spike times.
• Validated model predictions against simulation and analytical results.

Conclusions:

• The proposed stochastic models effectively capture neuron firing dynamics under time-varying input.
• The study offers a robust method for predicting successive spike times.
• The findings contribute to a deeper understanding of neuronal excitability and signal processing in the brain.