Deep-learning-assisted detection and termination of spiral and broken-spiral waves in mathematical models for cardiac tissue

Abstract

This paper proposes a low-amplitude defibrillation scheme, guided by the convolution neural network, for the removal of such waves in the mathematical models for cardiac tissue. The advantage of the scheme is that a snapshot of the spatial information about the membrane potential is sufficient to locate the source of the electrical activations, and further the application of low-amplitude Gaussian currents on these locations eliminates such waves.