This project presents a suite of advanced scientific computing tools for the analysis of Atrial Fibrillation (AF), focusing on the reconstruction of cardiac electrical activity starting from signals acquired via mapping catheters.
The project addresses the challenge of interpreting complex cardiac electrical signals during atrial fibrillation episodes. Using data from a star-shaped catheter equipped with 20 sensors, the study implements and compares different mathematical methodologies to reconstruct the electrical impulse propagation:
- Signal Analysis: Processing of local electrical potentials to identify activation times and classify the proximity to a focal site activity.
- Physics-Informed Neural Networks (PINNs): Exploration of Deep Learning techniques that integrate physical laws to reconstruct electrical impulse propagation. This approach balances underlying physical principles with sparse measurements, addressing the inverse problem in a natural way.
- Evolutionary Optimization and Fast Marching Method (FMM): Utilization of a physics-based approach to reconstruct the activation time field on the cardiac surface by solving the Eikonal equation. This is achieved through a surrogate computational model and evolutionary optimization to pinpoint activation centers.
- Uncertainty Quantification: Assessment of the method's stability by simulating sensor failure (disabling two sensors). This involves stochastic simulation techniques, sampling from parametric distributions of the estimated parameters, and collecting the outputs for statistical analysis.
The results demonstrate how the integration of physical models and optimization algorithms allows for robust localization of arrhythmia drivers, even in the presence of incomplete data.
The source code developed for this project is hosted in a private repository and cannot be made public due to privacy reasons. The report included here documents the entire workflow and results of the project.
The scientific report included in this repository is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0). You are free to share and redistribute the material in any medium or format, provided that appropriate credit is given.
For questions, clarifications, or further information about the project, feel free to contact me at mattia.gastoldi@mail.polimi.it. Upon request, further details can be provided.