Analytical Modelling of Surface EMG Signals Generated by Curvilinear Fibers With Approximate Conductivity Tensor

Area of Science:

• Biomedical Engineering
• Computational Neuroscience
• Electrophysiology

Background:

• Mathematical modeling of electromyographic (EMG) signals aids in interpreting experimental data and validating signal processing techniques.
• Existing analytical EMG models often simplify muscle fiber arrangements, neglecting variations in orientation along fiber paths.
• Accurate modeling requires accounting for complex, non-uniform muscle fiber trajectories.

Purpose of the Study:

• To develop a subject-specific analytical model for simulating EMG signals.
• To incorporate realistic, curvilinear muscle fiber trajectories in the upper arm.
• To analytically derive action potentials using an approximate conductivity tensor.

Main Methods:

• Magnetic Resonance (MR) imaging was used to reconstruct individual muscle fiber paths and build the volume conductor model.
• Action potentials were simulated along identified curvilinear fibers with an assumed cylindrically anisotropic conductivity tensor.
• Single fiber action potentials (SFAPs) were computed by modeling membrane current sources, with validation against numerical models using approximate and exact conductivity tensors.

Main Results:

• The proposed analytical model produced motor unit action potentials highly similar to numerical models (cross-correlation 0.98, nRMSE ≤ 0.04).
• Simulated compound muscle action potentials (CMAPs) closely matched experimentally recorded data.
• The model served as a realistic benchmark for evaluating EMG decomposition algorithms.

Conclusions:

• The developed analytical model accurately generates action potentials that represent the spatial distribution of muscle fibers with curvilinear paths.
• This subject-specific approach enhances the fidelity of EMG signal modeling.
• The model provides a valuable tool for validating EMG signal processing techniques and algorithms.