Evaluation of a Rapid Anisotropic Model for ECG Simulation.

Area of Science:

• Computational electrophysiology
• Medical imaging

Background:

• Complex cardiac electrophysiology models require high-performance computing, limiting clinical use.
• Current ECG imaging tools often use simplified models neglecting anisotropic conductivity and confining results to the heart-torso interface.

Purpose of the Study:

• To develop a computationally efficient forward model for cardiac electrophysiology that incorporates anisotropic tissue conductivity.
• To generate standard 12-lead electrocardiograms (ECGs) and activation maps rapidly for potential clinical integration.

Main Methods:

• Utilized an eikonal model for 3D myocardial activation sequence approximation.
• Employed the lead-field approach for ECG computation.
• Implemented both solvers on massively parallelized graphics processing units (GPUs) for speed and scalability.

Main Results:

• The proposed model accurately approximates activation maps and ECGs compared to the bidomain model.
• The computational approach achieved results in seconds, demonstrating significant speed improvement.
• Solvers exhibited excellent scalability on high-end hardware, confirming performance potential.

Conclusions:

• The developed methods provide a viable and fast alternative for cardiac electrophysiology modeling.
• These advancements are suitable for enhancing ECG imaging techniques.
• The speed of the solvers suggests potential for future interactive simulation tools in clinical settings.