Background: The heart's contraction is caused by electrical excitation, and it was recently shown in a simulated slab of myocardial muscle tissue that the electrical excitation can be computed from the motion of the tissue using deep learning.
Aims: Here, we demonstrate in computer simulations that it is possible to predict three-dimensional electrical wave dynamics from ventricular deformation mechanics using deep learning.
Methods: We performed thousands of simulations of focal and reentrant electromechanical activation dynamics in idealized bi-ventricular geometries, and used the data to train a neural network to analyze the ventricular deformation mechanics and subsequently predict the three-dimensional electrical wave pattern that caused the deformation.
Results: We demonstrate that, next to focal wave patterns, even complicated three-dimensional electrical scroll wave patterns can be reconstructed, even if the network has never seen the particular arrhythmia or heart geometry and was trained on a different electrophysiological model or with different mechanical properties. We show that the deep learning model has the ability to generalize by training it on data generated with the smoothed particle hydrodynamics (SPH) method and subsequently applying it to data generated with the finite element method (FEM). Predictions can be performed in the presence of scars and with significant heterogeneity.
Conclusion: Our results suggest that, with adequate training data, deep neural networks could be used to calculate intramural action potential wave patterns from imaging data of the motion of the heart muscle.