An admissible solution approach to inverse electrocardiography

Area of Science:

• Biomedical Engineering
• Computational Biology
• Medical Physics

Background:

• The inverse problem of electrocardiography aims to identify cardiac electrical activity from body surface potentials.
• Torso volume conduction effects render this problem ill-posed, requiring regularization in standard solutions.
• Existing methods may introduce bias due to chosen constraints and regularization parameters.

Purpose of the Study:

• To develop a novel approach for the inverse problem of electrocardiography by reformulating it as a search within an admissible solution set.
• To overcome limitations of standard regularized solutions, particularly their difficulty in incorporating multiple constraints and potential bias.
• To introduce a method that does not require a regularization parameter, instead focusing on defining constraint sets.

Main Methods:

• Reformulating the inverse problem as finding a solution within an admissible set defined by constraints.
• Utilizing multiple types of constraints, including spatial, weighted, and temporal.
• Employing an iterative convex optimization algorithm, specifically the ellipsoid algorithm.

Main Results:

• Demonstrated accuracy and feasibility of the proposed method through simulation results.
• Successfully applied the method using dipole sources and measured epicardial potentials.
• The approach allows for the integration of multiple constraints, unlike traditional techniques.

Conclusions:

• The developed method offers a robust and flexible approach to solving the inverse problem of electrocardiography.
• By defining admissible solution sets and utilizing multiple constraints, this technique enhances the reliability and accuracy of cardiac electrical activity characterization.
• The absence of a regularization parameter and the use of iterative convex optimization present a significant advancement in the field.