Unified Neural Network Based Pathologic Event Reconstruction Using Spatial Heart Model

Abstract

This paper presents a new way to solve the inverse problem of electrocardiography in terms of heart model parameters. The developed event estimation and recognition method uses a unified neural network (UNN)-based optimization system to determine the most relevant heart model parameters. A UNN-based preliminary ECG analyzer system has been created to reduce the searching space of the optimization algorithm. The optimal model parameters were determined by a relation between objective function minimization and robustness of the solution. The final evaluation results, validated by physicians, were about 96% correct. Starting from the fact that input ECGs contained various malfunction cases, such as Wolff-Parkinson-White (WPW) syndrome, atrial and ventricular fibrillation, these results suggest this approach provides a robust inverse solution, circumventing most of the difficulties of the ECG inverse problem.

Encouraging an appropriate representation simplifies training of neural networks

“…Although the appropriate representation depends on the underlying task, we believe that the general idea of encouraging models to learn concepts that are similar to human concepts may be beneficial in various applications ranging from the analysis of biomedical signals [10], such as ECG [17] to image classification [11]. Furthermore, in many cases, see e.g.…”