Source localization for gastric electrical activity using simulated magnetogastrographic data.

Area of Science:

• Biomedical Engineering
• Computational Physiology
• Medical Imaging

Background:

• Magnetogastrography (MGG) measures magnetic fields from the stomach.
• Understanding stomach slow wave (SW) propagation is crucial for diagnosing gastrointestinal disorders.
• Accurate source localization of SWs from MGG data remains a challenge.

Purpose of the Study:

• To investigate the efficacy of the magnetic dipole (MDP) approximation for localizing SW sources in MGG data.
• To assess the accuracy of MDP fitting in identifying SW propagation patterns.
• To evaluate the robustness of the MDP method under noisy conditions.

Main Methods:

• Simulated MGG data using an anatomically realistic torso and stomach model.
• Grid-based finite element method to model SW activity and compute magnetic fields (MFs).
• MDP fitting applied to time-sampled MFs, calculating goodness of fit (GOF) and distance to center of gravity (COG).

Main Results:

• MDP fitting successfully identified SW propagation patterns with high correlation scores (0.63-0.81 across directions).
• The mean distance between fitted MDPs and COGs of active dipoles was 49±4 mm.
• The method demonstrated similar performance under simulated Gaussian noise conditions.

Conclusions:

• MDP approximation is a viable tool for source localization in MGG data.
• This technique can accurately characterize SW propagation, offering insights into gastric electrical activity.
• The findings support the use of MDP for analyzing MGG signals in clinical and research settings.