Stochastic resonance in a sinusoidally forced LIF model with noisy threshold

Area of Science:

• Computational Neuroscience
• Nonlinear Dynamics
• Signal Processing

Background:

• The Leaky Integrate-and-Fire (LIF) model is a fundamental tool for simulating neuronal activity.
• Understanding the impact of noise on neural firing dynamics is crucial for deciphering brain function.
• Stochastic Resonance (SR) describes a phenomenon where noise can enhance signal detection in nonlinear systems.

Purpose of the Study:

• To investigate Stochastic Resonance (SR) in a LIF neural model with a Gaussian-distributed noise on the threshold.
• To analytically derive statistical functions describing neural firing activity.
• To explore the emergence of bimodal resonance curves and their underlying mechanisms.

Main Methods:

• Approximation of the LIF neural model using an instantaneous firing rate dependent on voltage.
• Analytical solution of the governing equations to obtain firing phase and interspike interval distributions.
• Analysis of derived quantities to demonstrate SR and characterize resonance curves.

Main Results:

• The study successfully derived key statistical functions for neural firing activity.
• Stochastic Resonance (SR) was demonstrated, showing noise-enhanced signal transmission.
• Bimodal resonance curves were observed under specific conditions (low frequencies/large amplitudes or high frequencies).

Conclusions:

• The LIF model with threshold noise exhibits both regular and bimodal Stochastic Resonance (SR).
• The bimodal feature of the resonance curves is attributed to phase-locked firing patterns.
• This research provides insights into noise-induced signal processing in neural systems.