Investigating Deep Learning Benchmarks for Electrocardiography Signal Processing
This work addresses the problem of fragmented and non-reproducible deep learning methods for ECG signal processing, providing a standardized tool for researchers, though it is incremental as it consolidates existing and novel approaches rather than introducing a new paradigm.
The authors tackled the lack of systematic studies and open-source libraries for deep learning in electrocardiography (ECG) processing by proposing torch_ecg, a framework that gathers neural networks and provides benchmark studies, resulting in a tool that offers convenient and modular building of networks and uniform data preparation for the ECG research community.
In recent years, deep learning has witnessed its blossom in the field of Electrocardiography (ECG) processing, outperforming traditional signal processing methods in various tasks, for example, classification, QRS detection, wave delineation. Although many neural architectures have been proposed in the literature, there is a lack of systematic studies and open-source libraries for ECG deep learning. In this paper, we propose a deep learning framework, named \texttt{torch\_ecg}, which gathers a large number of neural networks, both from literature and novel, for various ECG processing tasks. It establishes a convenient and modular way for automatic building and flexible scaling of the networks, as well as a neat and uniform way of organizing the preprocessing procedures and augmentation techniques for preparing the input data for the models. Besides, \texttt{torch\_ecg} provides benchmark studies using the latest databases, illustrating the principles and pipelines for solving ECG processing tasks and reproducing results from the literature. \texttt{torch\_ecg} offers the ECG research community a powerful tool meeting the growing demand for the application of deep learning techniques.