Dynamical Timescale Explains Marginal Stability in Excitability Dynamics.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Dynamical Systems

Background:

• Neural computation relies on action potentials occurring over milliseconds.
• Longer-term neural excitability dynamics, crucial for behavior, are not well understood.
• Previous studies observed fluctuating neuronal responses to constant stimuli and reliable responses to variable stimuli.

Purpose of the Study:

• To resolve the apparent paradox of neuronal responses to stimuli of varying statistics.
• To investigate the dynamic regimes governing neural excitability over behaviorally relevant timescales.
• To develop a predictive model for neural excitability dynamics.

Main Methods:

• Employed a novel inference method to identify the dynamic regime of neuronal excitability.
• Developed a new computational model incorporating a dynamic recovery timescale.
• Validated the model against experimental data under various stimulus conditions.

Main Results:

• Demonstrated that neurons operate in a marginally stable dynamic regime.
• Showed this regime is characterized by large internal fluctuations and high sensitivity to external stimuli.
• The novel model accurately predicts neuronal responses, explaining experimental observations with high fidelity.

Conclusions:

• Marginal stability is key to understanding neural excitability's dual response characteristics.
• A dynamic recovery timescale is a critical factor in modeling neuronal dynamics.
• The proposed model offers a parsimonious and powerful framework for studying neural excitability and network function.