An Adaptive Generalized Leaky Integrate-and-Fire Model for Hippocampal CA1 Pyramidal Neurons and Interneurons.

Area of Science:

• Computational Neuroscience
• Biophysics
• Systems Neuroscience

Background:

• Realistic neural network models are crucial for in-silico experiments and digital brain twins.
• Current simplified neuron models, like generalized leaky integrate-and-fire, lack the accuracy to replicate complex firing dynamics in regions like the hippocampus.
• Technical limitations in supercomputing necessitate efficient yet accurate neuron models.

Purpose of the Study:

• To develop an adaptive generalized leaky integrate-and-fire model capable of reproducing the complex firing dynamics of hippocampal CA1 neurons and interneurons.
• To enhance the accuracy of computational models for simulating neural activity in the hippocampus.
• To create a more computationally efficient yet biophysically realistic neuron model.

Main Methods:

• Proposed an adaptive generalized leaky integrate-and-fire model using linear ordinary differential equations with nonlinear initial and update conditions.
• Employed mathematical analysis of equilibria stability and membrane potential monotonicity to derive model constraints.
• Validated the model against experimental data from 85 hippocampal neurons and interneurons using various stimulation protocols.

Main Results:

• The adaptive model successfully reproduces the nonlinear firing dynamics of hippocampal neurons and interneurons.
• Mathematical analysis provided constraints that reduced computational cost for parameter optimization.
• The model quantitatively reproduces and predicts experimental traces for diverse stimulation protocols.
• Generated statistically indistinguishable synthetic neuron copies reflecting experimental variability.

Conclusions:

• The proposed adaptive neuron model offers a significant improvement over existing models for simulating hippocampal function.
• This approach enables the creation of large-scale, realistic neural network simulations with controlled firing properties.
• Facilitates more accurate in-silico investigations of cognitive functions and the development of digital brain twins.