Analytical integrate-and-fire neuron models with conductance-based dynamics and realistic postsynaptic potential time...

Area of Science:

• Computational neuroscience
• Neuronal modeling
• Computational performance

Background:

• Previous gIF neuron models offered analytical approximations for event-driven simulations.
• These models used simplified, non-realistic postsynaptic potential (PSP) time courses.
• Limitations included discontinuous PSPs, lacking realistic decay dynamics.

Purpose of the Study:

• To develop an analytical integrate-and-fire (IF) neuron model with a full postsynaptic potential (PSP) time course.
• To improve the biological realism of conductance-based (COBA) IF neuron models.
• To enhance computational performance in large-scale neural network simulations.

Main Methods:

• Developed an analytical approximation for the conductance-based (COBA) integrate-and-fire (IF) neuron model.
• Incorporated a full PSP time course, addressing limitations of previous models.
• Designed a computationally efficient algorithm for large-scale simulations.

Main Results:

• The new gIF4 model accurately reproduces subthreshold and suprathreshold responses compared to numerical solutions.
• Achieved a computational performance increase of at least two orders of magnitude.
• Demonstrated effective simulation of passive membrane equations with conductance noise.

Conclusions:

• The gIF4 model provides a computationally efficient and biologically realistic alternative for simulating neuronal networks.
• It overcomes the limitations of simplified PSP time courses in previous analytical models.
• The proposed implementation facilitates large-scale neural network simulations with enhanced accuracy and speed.