Made-to-order spiking neuron model equipped with a multi-timescale adaptive threshold

Area of Science:

• Computational Neuroscience
• Neuroscience
• Computational Biology

Background:

• Accurate neuronal modeling is crucial for understanding brain function and cognitive processes.
• Existing quantitative models struggle with prediction accuracy and high computational demands.
• Bridging the gap between qualitative and quantitative neuronal modeling is essential.

Purpose of the Study:

• To develop a simple, fast computational model for accurately predicting diverse neuronal spike responses.
• To create a model adaptable to various cortical neuron types.
• To facilitate faithful modeling of real brain function and network properties.

Main Methods:

• Devised a computational model featuring a multi-timescale adaptive threshold predictor and a nonresetting leaky integrator.
• Utilized three adaptive threshold parameters to reproduce various neuronal firing patterns.
• Employed a three-dimensional parameter space to represent continuous firing characteristics.

Main Results:

• The model successfully reproduces diverse neuronal spike responses, including regular spiking, intrinsic bursting, fast spiking, and chattering.
• Achieved accurate prediction of neuronal responses to fluctuating currents.
• Demonstrated flexibility in capturing a continuous spectrum of firing characteristics beyond discrete categories.

Conclusions:

• The proposed model offers high flexibility and low computational cost for neuronal simulation.
• Enables faithful modeling of real brain function by accounting for distributed neuronal characteristics.
• Facilitates the examination of network properties influenced by individual neuron variability.