Bayesian approach for localizing cardiac sources in Magnetocardiography using Vectorcardiography based total...

Area of Science:

• Biophysics
• Biomedical Engineering
• Cardiology

Background:

• The human heart generates electrical signals for muscle contraction, creating measurable surface potentials and magnetic fields.
• Investigating these signals via Electro/Magnetocardiogram (E/MCG) is crucial for understanding cardiac function.
• A significant challenge in E/MCG is the 3D source localization of cardiac dysfunctions, known as the inverse problem.

Purpose of the Study:

• To develop and validate a novel algorithm for improved cardiac source localization using Magnetocardiography (MCG).
• To integrate Vectocardiography (VCG) signals into the forward and inverse problems of MCG for enhanced diagnostic accuracy.
• To compare the efficacy of Bayesian and deterministic approaches for solving ill-posed inverse problems in MCG, particularly in the context of Myocardial Ischemia.

Main Methods:

• A homogeneous volume conductor model was employed for the forward problem construction in MCG.
• A novel algorithm was implemented utilizing Vectocardiography (VCG) signals within the MCG forward problem.
• Dynamic lead field, based on VCG orientations, was used for the inverse problem.
• Bayesian and deterministic approaches were utilized to solve ill-posed inverse problems, with results analyzed for noise signal measurements.

Main Results:

• The proposed algorithms, incorporating VCG priors within a Bayesian framework, significantly enhanced MCG source localization accuracy for Myocardial Ischemia.
• Bayesian and total variation methods, utilizing VCG priors, reduced the spread of reconstructed inverse solution borders to 2.9-3.03cm, compared to 3.5-4.3cm for deterministic methods.
• The integration of VCG signals improved the accuracy of cardiomagnetic imaging and demonstrated potential for more reliable cardiac diagnostic tools.

Conclusions:

• The incorporation of VCG signals into the MCG forward problem enhances the precision of cardiomagnetic imaging.
• The developed Bayesian approach with VCG priors offers a more reliable method for diagnosing cardiac dysfunctions, particularly Myocardial Ischemia.
• This research paves the way for advanced, reliable diagnostic tools in clinical cardiology.