Reconstruction of multiple gastric electrical wave fronts using potential-based inverse methods.

Area of Science:

• Biomedical Engineering
• Computational Physiology
• Medical Imaging

Background:

• Non-invasive characterization of gastric electrical activity is crucial for understanding gastrointestinal function.
• Solving the inverse problem, reconstructing stomach surface potentials from skin measurements, is a key challenge.
• Existing methods require validation in realistic anatomical models.

Purpose of the Study:

• To evaluate the Greensite-Tikhonov and Tikhonov inverse methods for reconstructing gastric electrical activity.
• To assess the impact of electrode number and noise on reconstruction accuracy.
• To determine the stability of the Greensite-Tikhonov method under varying conditions.

Main Methods:

• Anatomically realistic torso and stomach models were used for simulations.
• Gastric electrical activity was simulated, and potentials were recorded on stomach and skin surfaces.
• Greensite-Tikhonov and Tikhonov methods were compared using 84 or 204 body surface electrodes with 10% Gaussian noise.

Main Results:

• The Greensite-Tikhonov method successfully represented propagating wave fronts.
• The Greensite-Tikhonov method with 204 electrodes achieved 90% activation time correlation and 3 cm localization error.
• This method remained stable with up to 20% Gaussian noise and 10% stomach size variation.

Conclusions:

• The Greensite-Tikhonov method is a stable and accurate approach for non-invasive gastric electrical activity reconstruction.
• Increasing the number of body surface electrodes enhances the performance of the Greensite-Tikhonov method.
• Greensite-Tikhonov consistently outperformed the Tikhonov method in all tested scenarios.