Accurate simulation of cuff electrode stimulation predicting in-vivo strength-duration thresholds

Area of Science:

• Biomedical Engineering
• Computational Neuroscience

Background:

• In-silico experiments can optimize peripheral nerve electrode designs, reducing animal use and informing novel stimulation techniques.
• Current in-silico models often lack quantitative agreement with in-vivo results, suggesting environmental realism is crucial.
• Nerve bundle lamination and extraneural environment significantly impact simulation accuracy.

Purpose of the Study:

• To assess the sensitivity of in-silico nerve models to electrical parameter estimates and volume conductor types.
• To improve the quantitative accuracy of in-silico experiments for peripheral nerve electrode design.
• To compare in-silico predictions with in-vivo experimental data for rat vagus nerves.

Main Methods:

• In-vivo strength-duration curves were measured on rat vagus nerves using a needle electrode.
• An in-silico analog model was created using finite element method (FEM) simulations.
• Simulations varied extraneural environment (in-saline vs. in-air), tissue resistivity (ρ), and permittivity (εr).
• FEM results were projected onto a McIntyre, Richardson, and Grill (MRG) nerve fiber model in NEURON.

Main Results:

• The most accurate in-silico model used an in-air environment with low-frequency resistivity and measured permittivity.
• This specific combination achieved high convergence (r² = 0.96) with in-vivo data.
• In-air boundary conditions in simulations closely matched the in-vivo experimental setup.

Conclusions:

• Increased simulation realism, particularly extraneural environment and resistivity, leads to more accurate predictions.
• Electrical parameters are critical for accurate modeling of nerve stimulation and recording.
• Reactive electrical parameters become important for high-frequency waveforms.