Fast simulation of extracellular action potential signatures based on a morphological filtering approximation

Area of Science:

• Computational Neuroscience
• Neurophysiology
• Signal Processing

Background:

• Simulating extracellular recordings of neuronal populations is crucial for understanding neural dynamics and validating signal analysis tools.
• Detailed multicompartmental neuron models (e.g., NEURON) offer realism but incur significant computational costs, limiting large-scale simulations.
• Existing methods face challenges in balancing accuracy with computational efficiency for simulating large neuronal populations.

Purpose of the Study:

• To develop a novel, computationally efficient method for simulating extracellular potentials of firing neurons.
• To validate the proposed method across different neuron morphologies and assess its ability to reproduce realistic extracellular action potentials.

Main Methods:

• Developed a linear geometry-based filter to model action potential generation and propagation along the neuron's axon.
• Incorporated neuron geometry and electrode positions into the simulation.
• Assessed the validity of the approach for various neuron morphologies, focusing on the axon/dendrites surface ratio.

Main Results:

• The proposed method successfully simulates realistic extracellular action potentials.
• The approach demonstrates computational efficiency, making large-scale simulations feasible.
• The method's validity was confirmed across a range of neuron morphologies.

Conclusions:

• The novel linear filter-based method provides a time-efficient alternative for simulating extracellular neuronal recordings.
• This approach facilitates the study of extracellular field potentials and the validation of spike detection and sorting algorithms.
• The method offers a practical solution for researchers needing to simulate large neuronal populations without high computational burden.